Existing paper review methods often rely on superficial manuscript features or directly on large language models (LLMs), which are prone to hallucinations, biased scoring, and limited reasoning capabilities. Moreover, these methods often fail to capture the complex argumentative reasoning and negotiation dynamics inherent in reviewer-author interactions. To address these limitations, we propose ReViewGraph (Reviewer-Author Debates Graph Reasoner), a novel framework that performs heterogeneous graph reasoning over LLM-simulated multi-round reviewer-author debates. In our approach, reviewer-author exchanges are simulated through LLM-based multi-agent collaboration. Diverse opinion relations (e.g., acceptance, rejection, clarification, and compromise) are then explicitly extracted and encoded as typed edges within a heterogeneous interaction graph. By applying graph neural networks to reason over these structured debate graphs, ReViewGraph captures fine-grained argumentative dynamics and enables more informed review decisions. Extensive experiments on three datasets demonstrate that ReViewGraph outperforms strong baselines with an average relative improvement of 15.73%, underscoring the value of modeling detailed reviewer–author debate structures.

Xin Chen, Feng Jiang, Yiqian Zhang, Hardy Chen, Shuo Yan, Wenya Xie, Min Yang, Shujian Huang. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Feiteng Fang, Dingwei Chen, Xiang Huang, Ting-En Lin, Yuchuan Wu, Xiong Liu, Jing Ye, Ziqiang Liu, Haonan Zhang, Liang Zhu, Hamid Alinejad-Rokny, Min Yang, Yongbin Li. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Text-to-visualization (text-to-vis) models for tabular data have become essential tools in the era of big data, enabling users to generate visualizations and make data-driven decisions through natural language queries (NLQs). Despite their growing adoption, the security vulnerabilities of these models remain largely unexplored. To address this gap, we propose VisPoison, a backdoor attack framework that realistically simulates three types of attacks on text-to-vis models via data poisoning: data exposure, misleading visualizations, and denial-of-service (DoS). Specifically, VisPoison introduces two types of stealthy triggers to enable both proactive and passive backdoor activations. Proactive triggers are deliberately inserted by attackers using rare-word patterns to extract sensitive information, whereas passive triggers are unintentionally activated by users through first-word prompts, resulting in visualization errors or DoS failures. To support these triggers, we craft specialized payloads for visualization queries that allow compromised models to function normally on benign inputs while producing malicious outputs in the presence of triggers. Extensive evaluations on both trainable and in-context learning (ICL)-based text-to-vis models show that VisPoison achieves attack success rates exceeding 90\%, exposing serious vulnerabilities. Additionally, existing defense strategies reveal limited effectiveness against VisPoison, underscoring the urgent need for more robust and security-aware text-to-vis systems to safeguard human-data interaction.

E-commerce platforms increasingly rely on Large Language Models (LLMs) and Vision Language Models (VLMs) to detect illicit or misleading product content. However, these models remain vulnerable to evasive content, which refers to inputs that have been deliberately modified through techniques such as word splitting, euphemistic language, or image cropping to conceal policy violations while still conveying prohibited claims. Crucially, detecting such content requires a model to simultaneously master two capabilities: accurately comprehending complex rules, and correctly inferring the true intent behind deliberately obfuscated multimodal inputs. While prior work has separately explored LLM reasoning over complex rules and LLM-based detection of evasive content, no existing benchmark combines both within a unified evaluation framework. This gap is particularly consequential in e-commerce, where accurate moderation demands that both capabilities operate in concert. To address this gap, we introduce EVADE-Bench, the first expert-curated Chinese multimodal benchmark specifically designed to evaluate LLMs and VLMs on evasive content detection in real-world e-commerce scenarios. The dataset contains 2,833 annotated text samples and 13,961 annotated images spanning six violation categories. Our comprehensive evaluation of 26 open- and closed-source LLMs and VLMs reveals that even state-of-the-art models frequently misclassify evasive samples. We further demonstrate that clearer rule categorization significantly improves model prediction consistency and reduces false predictions, highlighting the critical role of benchmark design in enabling reliable evaluation. We analyze common error patterns across these models and identify systematic limitations in their ability to reason over metaphorical expressions and complex regulatory rules. To explore paths for performance improvement, we investigate the feasibility of multi-agent decomposition for multimodal reasoning, wherein visual description and logical inference are decoupled into separate agents, and find that this strategy yields notable accuracy gains. By releasing EVADE-Bench, we provide the first rigorous standard for evaluating evasive content detection and aim to support the development of safer and more trustworthy content moderation systems. The dataset is publicly available at https://huggingface.co/datasets/koenshen/EVADE-Bench.

Zero-shot single-cell cell-type annotation aims to determine a cell's type from a given set of expressed genes without any training. Existing knowledge-graph-based RAG approaches retrieve evidence by expanding from source entities and relying on iterative LLM reasoning. However, in this setting each query contains tens to hundreds of genes, where no single gene is decisive and the label emerges only from their collective co-occurrence. Such hyper-entity queries fundamentally challenge local, entity-wise exploration strategies, which reason from individual genes, leading to poor scalability and substantial LLM cost. We propose GATHER (Graph-Aware Traversal with Hyper-Entity Retrieval), a convergence-centric retriever tailored to hyper-entity queries. It performs global multi-source graph traversal and identifies topological convergence points, nodes jointly reachable from many input genes. These convergence nodes act as high-information hyper-entities that capture entity synergy. By incorporating node- and path-importance scoring, GATHER selects informative evidence entirely without LLM involvement during retrieval. Instantiated on a self-constructed cell-centric biological knowledge graph (VCKG), GATHER outperforms strong KG-RAG baselines (ToG, ToG-2, RoG, PoG) on two datasets (Immune and Lung), achieving the highest exact-match accuracy (27.45% and 59.64%) with only a single LLM call per sample, compared to 2 to 61 calls for KG-RAG baselines. Our results demonstrate that convergence nodes compress multi-entity signals into compact, high-information evidence that conveys more per item than multi-hop paths, providing an efficient global alternative to local entity-wise reasoning.

Abstract Stance detection on social media plays a vital role in understanding public opinion on contentious topics. While prior work leverages external knowledge sources like Wikipedia to enrich limited target information, it primarily introduces conceptual content, neglecting the interpretability potential of knowledge and often leading to the incorporation of irrelevant or redundant information that hinders stance prediction performance. To address this, we introduce PiKGL, a Pruned interpretable Knowledge Graph Learning framework for explainable stance detection. Specifically, we first extract event triplets and topics to obtain real-world knowledge, which is then used to construct an interpretable knowledge graph. To ensure precision and minimize noise, we introduce a retrieval-guided pruning strategy that incorporates commonsense knowledge, filtering redundant information of the interpretable knowledge graph. Finally, the pruned knowledge graph is injected into a large language model to jointly model textual, target, and commonsense for improved stance comprehension. Experimental results conducted on three public datasets demonstrate our PiKGL achieves state-of-the-art performance on stance detection.

The long-tail recommendation problem remains a significant challenge in modern recommender systems, primarily due to data sparsity and popularity bias, which hinder the accurate ID representation of users and items. Recent advancements in large language models (LLMs) have enabled the direct modeling of user and item semantic representations, offering potential improvements in representation learning through the alignment of these two types of representations. However, systems relying on LLM representation alignment face two critical challenges: (1) the substantial differences between LLMs and recommendation models in terms of training objectives, phases, and data; (2) the pervasive popularity bias in collaborative data. These challenges create a semantic gap between ID representations and semantic representations. Directly aligning these representations risks introducing recommendation-irrelevant noise, disrupting the collaborative information embedded in ID representations, and ultimately leading to suboptimal recommendation outcomes. To address this gap, we propose DeltaRec, aDouble-enhancement framework forlong-tailRecommendation. DeltaRec tackles the long-tail recommendation problem through two approaches. First, it incorporates semantic information for all items. Second, it provides additional supervision signals specifically for long-tail items. The framework begins by disentangling ID representations into interest representations and conformity representations. To integrate semantic information from LLMs while preserving popularity information, we design a contrastive learning-based semantic alignment module that aligns interest representations with semantic representations. Furthermore, to enhance the representation learning of unpopular items, we introduce a ranking-based behavior alignment module, which provides additional supervision signals for these items. To avoid introducing recommendation-irrelevant noise and disrupting collaborative semantics due to excessive alignment, we propose a curriculum learning-based training mechanism. Extensive experiments on real-world datasets demonstrate that DeltaRec effectively mitigates popularity bias and significantly improves long-tail recommendation performance without relying on prior knowledge of popularity distributions. Our code is available athttps://github.com/leo0481/DeltaRec/E3D7.

The success of Large Language Models (LLMs) relies heavily on the huge amount of pre-training data learned in the pre-training phase. The opacity of the pre-training process and the training data causes the results of many benchmark tests to become unreliable. If any model has been trained on a benchmark test set, it can seriously hinder the health of the field. In order to automate and efficiently test the capabilities of large language models, numerous mainstream benchmarks adopt a multiple-choice format. As the swapping of the contents of multiple-choice options does not affect the meaning of the question itself, we propose a simple and effective data leakage detection method based on this property. Specifically, we shuffle the contents of the options in the data to generate the corresponding derived data sets, and then detect data leakage based on the model's log probability distribution over the derived data sets. If there is a maximum and outlier in the set of log probabilities, it indicates that the data is leaked. Our method is able to work under gray-box conditions without access to model training data or weights, effectively identifying data leakage from benchmark test sets in model pre-training data, including both normal scenarios and complex scenarios where options may have been shuffled intentionally or unintentionally. Through experiments based on two LLMs and benchmark designs, we demonstrate the effectiveness of our method. In addition, we evaluate the degree of data leakage of 35 mainstream open-source LLMs on four benchmark datasets and give a ranking of the leaked LLMs for each benchmark, and we find that the Qwen family of LLMs has the highest degree of data leakage.

Stickers are widely used in online chatting, which can vividly express someone's intention, emotion, or attitude. Existing conversation research typically retrieves stickers based on a single session or the previous textual information, which can not adapt to the multi-modal and multi-session nature of the real-world conversation. To this end, we introduce MultiChat, a new dataset for sticker retrieval facing the multi-modal and multi-session conversation, comprising 1,542 sessions, featuring 50,192 utterances and 2,182 stickers. Based on the created dataset, we propose a novel Intent-Guided Sticker Retrieval (IGSR) framework that retrieves stickers for multi-modal and multi-session conversation history drawing support from intent learning. Specifically, we introduce sticker attributes to better leverage the sticker information in multi-modal conversation, which are incorporated with utterances to construct a memory bank. Further, we extract relevant memories for the current conversation from the memory bank to identify the intent of the current conversation, and then retrieve a sticker to respond guided by the intent. Extensive experiments on our MultiChat dataset reveal the robustness and effectiveness of our IGSR approach in multi-session, multi-modal scenarios.

Some of the latest released Code Large Language Models (Code LLMs) have been trained on repository-level code data, enabling them to perceive repository structures and utilize cross-file code information. This capability allows us to directly concatenate the content of repository code files in prompts to achieve repository-level code completion. However, in real development scenarios, directly concatenating all code repository files in a prompt can easily exceed the context window of Code LLMs, leading to a significant decline in completion performance. Additionally, overly long prompts can increase completion latency, negatively impacting the user experience. In this study, we conducted extensive experiments, including completion error analysis, topology dependency analysis, and cross-file content analysis, to investigate the factors affecting repository-level code completion. Based on the conclusions drawn from these preliminary experiments, we proposed a strategy called **Hierarchical Context Pruning (HCP)** to construct high-quality completion prompts. We applied the **HCP** to six Code LLMs and evaluated them on the CrossCodeEval dataset. The experimental results showed that, compared to previous methods, the prompts constructed using our **HCP** strategy achieved higher completion accuracy on five out of six Code LLMs. Additionally, the **HCP** managed to keep the prompt length around 8k tokens (whereas the full repository code is approximately 50k tokens), significantly improving completion throughput. Our code and data will be publicly available.

As the capabilities of Multimodal Large Language Models (MLLMs) improve, the need for higher-order evaluation of them is increasing. However, there is a lack of work evaluating MLLM for higher-order perception and understanding of Chinese visual content. To address this, we introduce the CII-Bench, which aims to assess MLLMs’ such capabilities for Chinese images. To ensure the authenticity of the Chinese context, images in CII-Bench are sourced from the Chinese Internet and manually reviewed, with corresponding answers also manually crafted. Additionally, CII-Bench incorporates images that represent Chinese traditional culture, such as famous Chinese traditional paintings, which can deeply reflect the model’s understanding of Chinese traditional culture. Through experiments on multiple MLLMs using CII-Bench, significant findings emerged. There is a large gap between MLLMs and humans in performance. The highest MLLM accuracy is 64.4%, while the human average is 78.2% and the peak is 81.0%. MLLMs perform poorly on traditional culture images, indicating limitations in understanding high-level semantics and lacking a deep knowledge base of Chinese traditional culture. Moreover, most models have higher accuracy when image emotion hints are added to the prompts. We believe CII-Bench will help MLLMs better understand Chinese semantics and specific images, and move forward the development of expert artificial general intelligence (AGI). Our project is publicly available at https://cii-bench.github.io.

Longze Chen, Renke Shan, Huiming Wang, Lu Wang, Ziqiang Liu, Run Luo, Jiawei Wang, Hamid Alinejad-Rokny, Min Yang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Sunbowen Lee, Junting Zhou, Chang Ao, Kaige Li, Xeron Du, Sirui He, Haihong Wu, Tianci Liu, Jiaheng Liu, Hamid Alinejad-Rokny, Min Yang, Yitao Liang, Zhoufutu Wen, Shiwen Ni. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Yang Sun, Guanrong Chen, Hamid Alinejad-Rokny, Jianzhu Bao, Yuqi Huang, Bin Liang, Kam-Fai Wong, Min Yang, Ruifeng Xu. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Haonan Zhang, Run Luo, Xiong Liu, Yuchuan Wu, Ting-En Lin, Pengpeng Zeng, Qiang Qu, Feiteng Fang, Min Yang, Lianli Gao, Jingkuan Song, Fei Huang, Yongbin Li. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Numerous medical systems powered by Large Language Models (LLMs) have achieved substantial progress, enabling them to perform diverse healthcare tasks. However, existing research is limited by the absence of real-world datasets specifically addressing medication safety, primarily due to privacy regulations and data accessibility challenges. Moreover, the evaluation of LLM-based systems in realistic clinical consultation settings, particularly with respect to medication safety, remains underexplored. To bridge these gaps, we propose a novel frame-work to simulate and evaluate clinical consultation scenarios for systematically assessing the medication safety capabilities of LLMs. Within this framework, we generate inquiry-diagnosis dialogues embedded with relevant medication risks and construct a dedicated medication safety database, RxRisk DB, comprising 6,725 contraindications, 28,781 drug interactions, and 14,906 indication-drug pairs. A two-stage filtering strategy ensures clinical realism and professional quality, resulting in the final benchmark, RxSafeBench, which includes 2,443 high-quality consultation scenarios evenly split across contraindication and interaction types. We evaluate state-of-the-art open-source and proprietary LLMs using a structured multiple-choice format that tests the models' ability to recommend the most appropriate medication given simulated patient context. Results reveal that current LLMs struggle to reliably incorporate contraindication and drug interaction information, especially when risks are implied rather than explicitly stated. Our findings highlight key challenges in deploying LLMs for medication safety and offer insights into improving their reliability through enhanced prompting strategies and task-specific fine-tuning. By introducing RxSafeBench, we provide the first comprehensive benchmark for assessing medication safety in LLMs, paving the way toward safer and more trustworthy AI-driven clinical decision support systems. Our code and data are released at https://github.com/CAS-SIAT-XinHai/RxSafeBench.

In this paper, we introduce a novel psychological benchmark, CPsyExam, constructed from questions sourced from Chinese examination systems. CPsyExam is designed to prioritize psychological knowledge and case analysis separately, recognizing the significance of applying psychological knowledge to real-world scenarios. We collect 22k questions from 39 psychology-related subjects across four Chinese examination systems. From the pool of 22k questions, we utilize 4k to create the benchmark that offers balanced coverage of subjects and incorporates a diverse range of case analysis techniques. Furthermore, we evaluate a range of existing large language models (LLMs), spanning from open-sourced to proprietary models. Our experiments and analysis demonstrate that CPsyExam serves as an effective benchmark for enhancing the understanding of psychology within LLMs and enables the comparison of LLMs across various granularities.

Dingwei Chen, Ziqiang Liu, Feiteng Fang, Chak Tou Leong, Shiwen Ni, Ahmadreza Argha, Hamid Alinejad-Rokny, Min Yang, Chengming Li. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

With the different roles that AI is expected to play in human life, imbuing large language models (LLMs) with different personalities has attracted increasing research interest.While the "personification" enhances human experiences of interactivity and adaptability of LLMs, it gives rise to critical concerns about content safety, particularly regarding bias, sentiment, and toxicity of LLM generation.This study explores how assigning different personality traits to LLMs affects the toxicity and biases of their outputs.Leveraging the widely accepted HEXACO personality framework developed in social psychology, we design experimentally sound prompts to test three LLMs' performance on three toxic and bias benchmarks.The findings demonstrate the sensitivity of all three models to HEXACO personality traits and, more importantly, a consistent variation in the biases, negative sentiment, and toxicity of their output.In particular, adjusting the levels of several personality traits can effectively reduce bias and toxicity in model performance, similar to humans' correlations between personality traits and toxic behaviors.The findings highlight the additional need to examine content safety besides the efficiency of training or fine-tuning methods for LLM personification, they also suggest a potential for the adjustment of personalities to be a simple and low-cost method to conduct controlled text generation.

Jing Xiong, Jianghan Shen, Fanghua Ye, Chaofan Tao, Zhongwei Wan, Jianqiao Lu, Xun Wu, Chuanyang Zheng, Zhijiang Guo, Min Yang, Lingpeng Kong, Ngai Wong. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Legal question answering (QA) has attracted increasing attention from people seeking legal advice, which aims to retrieve the most applicable answers from a large-scale database of question-answer pairs. Previous methods mainly use a dual-encoder architecture to learn dense representations of both questions and answers. However, these methods could suffer from lacking domain knowledge and sufficient labeled training data. In this paper, we propose a three-stage (pre-training, fine-tuning and re-ranking) framework for legal QA (called PFR-LQA), which promotes the fine-grained text representation learning and boosts the performance of dense retrieval with the dual-encoder architecture. Concretely, we first conduct domain-specific pre-training on legal questions and answers through a self-supervised training objective, allowing the pre-trained model to be adapted to the legal domain. Then, we perform task-specific fine-tuning of the dual-encoder on legal question-answer pairs by using the supervised learning objective, leading to a high-quality dual-encoder for the specific downstream QA task. Finally, we employ a contextual re-ranking objective to further refine the output representations of questions produced by the document encoder, which uses contextual similarity to increase the discrepancy between the anchor and hard negative samples for better question re-ranking. We conduct extensive experiments on a manually annotated legal QA dataset. Experimental results show that our PFR-LQA method achieves better performance than the strong competitors for legal question answering.

Image multi-emotion prediction aims to identify the emotions evoked by images in humans. In the real world, individual cognitive differences can lead to different viewers experiencing varied emotions. Most existing researchers primarily focus on analyzing image features, which are limited to the perceptual level, leading to a superficial understanding of emotions. To address this gap, we propose an Emotional Reasoning Chain (EReC) based on a multimodal large language model, which learns both perception and reasoning abilities for multi-emotion prediction. Specifically, we design a parameter-efficient fine-tuning paradigm encompassing three task instructions: perception, reasoning, and prediction instruction. This paradigm involves a progressive process to perform targeted instructions within a domain for fine-tuning, thereby optimizing the model’s capabilities in image perception, cognitive reasoning, and multi-emotion prediction. Furthermore, to alleviate hallucinations of large models, a Reason-level Alignment Score (RAS) is introduced to guide the model toward closer alignment with human cognition. Experiments on four datasets demonstrate that our EReC method, which emulates the human cognitive process of viewing and interpreting images, employs progressive tuning to integrate the perceptual, reasoning, and predictive capabilities of large models, achieving superior performance.

Federated learning (FL) enables collaborative training of a global model in the centralized server with data from multiple parties while preserving privacy. However, data heterogeneity can significantly degrade the performance of the global model when each party uses datasets from different sources to train a local model, thereby affecting personalized local models. Among various cases of data heterogeneity, feature drift, feature space difference among parties, is prevalent in real-life data but remains largely unexplored. Feature drift can distract feature extraction learning in clients and thus lead to poor feature extraction and classification performance. To tackle the problem of feature drift in FL, we propose FedPall, an FL framework that utilizes prototype-based adversarial learning to unify feature spaces and collaborative learning to reinforce class information within the features. Moreover, FedPall leverages mixed features generated from global prototypes and local features to enhance the global classifier with classification-relevant information from a global perspective. Evaluation results on three representative feature-drifted datasets demonstrate FedPall's consistently superior performance in classification with feature-drifted data in the FL scenario.

The development of large language models (LLMs) has significantly advanced the emergence of large multimodal models (LMMs). While LMMs have achieved tremendous success by promoting the synergy between multimodal comprehension and creation, they often face challenges when confronted with out-of-distribution data, such as which can hardly distinguish orientation, quantity, color, structure, etc. This is primarily due to their reliance on image encoders trained to encode images into task-relevant features, which may lead them to disregard irrelevant details. Delving into the modeling capabilities of diffusion models for images naturally prompts the question: Can diffusion models serve as the eyes of large language models for image perception? In this paper, we propose DEEM, a simple but effective approach that utilizes the generative feedback of diffusion models to align the semantic distributions of the image encoder. This addresses the drawbacks of previous methods that solely relied on image encoders like CLIP-ViT, thereby enhancing the model's resilience against out-of-distribution samples and reducing visual hallucinations. Importantly, this is achieved without requiring additional training modules and with fewer training parameters. We extensively evaluated DEEM on both our newly constructed RobustVQA benchmark and other well-known benchmarks, POPE and MMVP, for visual hallucination and perception. In particular, DEEM improves LMM's visual perception performance to a large extent (e.g., 4\% ↑ on RobustVQA, 6.5\% ↑ on MMVP and 12.8 \% ↑ on POPE ). Compared to the state-of-the-art interleaved content generation models, DEEM exhibits enhanced robustness and a superior capacity to alleviate model hallucinations while utilizing fewer trainable parameters, less pre-training data (10\%), and a smaller base model size. Extensive experiments demonstrate that DEEM enhances the performance of LMMs on various downstream tasks without inferior performance in the long term, including visual question answering, image captioning, and text-conditioned image synthesis.

Multimodal Emotion Recognition in Conversations (MERC) aims to detect emotions expressed in each utterance within conversational videos. Graph-based methods are widely employed in MERC due to their superiority in modeling intricate speaker-sensitive and context-sensitive dependencies in conversations. Despite promising advancements made, existing graph-based methods primarily suffer from two inherent issues due to their reliance on manually predefined graph structures: structural redundancy, which burdens models with irrelevant noise aggregation, and insufficient connections, which results in a lack of cross-modal contextual cues. To address the above issues, we propose a novel graph structure learning framework for MERC, which comprises two key components: Context-aware Graph Sparsification (CGS) and Implicit Graph Relation Mining (IGR). CGS employs an edge selection network to refine the manually predefined graph, filtering out noisy information caused by structural redundancy. IGR explores potential connections that are beneficial for emotional reasoning. Experimental results on two datasets show that our proposed framework significantly improves the performance of graph-based methods in MERC.

We introduce **SWE-Flow**, a novel data synthesis framework grounded in Test-Driven Development (TDD). Unlike existing software engineering data that rely on human-submitted issues, **SWE-Flow** automatically infers incremental development steps directly from unit tests, which inherently encapsulate high-level requirements. The core of **SWE-Flow** is the construction of a Runtime Dependency Graph (RDG), which precisely captures function interactions, enabling the generation of a structured, step-by-step *development schedule*. At each step, **SWE-Flow** produces a partial codebase, the corresponding unit tests, and the necessary code modifications, resulting in fully verifiable TDD tasks. With this approach, we generated 16,061 training instances and 2,020 test instances from real-world GitHub projects, creating the **SWE-Flow-Eval** benchmark. Our experiments show that fine-tuning open model on this dataset significantly improves performance in TDD-based coding. To facilitate further research, we release all code, datasets, models, and Docker images at [Github](https://github.com/Hambaobao/SWE-Flow).

Finite-sum Coupled Compositional Optimization (FCCO), characterized by its coupled compositional objective structure, emerges as an important optimization paradigm for addressing a wide range of machine learning problems. In this paper, we focus on a challenging class of non-convex non-smooth FCCO, where the outer functions are non-smooth weakly convex or convex and the inner functions are smooth or weakly convex. Existing state-of-the-art result face two key limitations: (1) a high iteration complexity of $O(1/\epsilon^6)$ under the assumption that the stochastic inner functions are Lipschitz continuous in expectation; (2) reliance on vanilla SGD-type updates, which are not suitable for deep learning applications. Our main contributions are two fold: (i) We propose stochastic momentum methods tailored for non-smooth FCCO that come with provable convergence guarantees; (ii) We establish a **new state-of-the-art** iteration complexity of $O(1/\epsilon^5)$. Moreover, we apply our algorithms to multiple inequality constrained non-convex optimization problems involving smooth or weakly convex functional inequality constraints. By optimizing a smoothed hinge penalty based formulation, we achieve a **new state-of-the-art** complexity of $O(1/\epsilon^5)$ for finding an (nearly) $\epsilon$-level KKT solution. Experiments on three tasks demonstrate the effectiveness of the proposed algorithms.

Recent advancements in omnimodal learning have significantly improved understanding and generation across images, text, and speech, yet these developments remain predominantly confined to proprietary models. The lack of high-quality omnimodal datasets and the challenges of real-time emotional speech synthesis have notably hindered progress in open-source research. To address these limitations, we introduce OpenOmni, a two-stage training framework that integrates omnimodal alignment and speech generation to develop a state-of-the-art omnimodal large language model. In the alignment phase, a pretrained speech model undergoes further training on image-text tasks, enabling (near) zero-shot generalization from vision to speech, outperforming models trained on tri-modal datasets. In the speech generation phase, a lightweight decoder is trained on speech tasks with direct preference optimization, which enables real-time emotional speech synthesis with high fidelity. Extensive experiments demonstrate that OpenOmni surpasses state-of-the-art models across omnimodal, vision-language, and speech-language benchmarks. It achieves a 4-point absolute improvement on OmniBench over the leading open-source model VITA, despite using 5$\times$ fewer training examples and a smaller model size (7B vs. 7$\times$8B). Besides, OpenOmni achieves real-time speech generation with less than 1 second latency at non-autoregressive mode, reducing inference time by 5$\times$ compared to autoregressive methods, and improves emotion classification accuracy by 7.7\%. The codebase is available at https://github.com/RainBowLuoCS/OpenOmni.

Large vision-language models (LVLMs) have emerged as foundational tools for real-world AI applications. Despite their remarkable capabilities, current LVLMs process entire images at the token level, leading to significant inefficiencies compared to human cognition, which selectively focuses on high-level vision concepts. This token-level redundancy becomes increasingly problematic for high-resolution images and long video sequences, resulting in large computational costs and limited scalability in practical applications. To address this limitation, we introduce the concept of a vision concept model, a novel paradigm that enables LVLMs to dynamically extract the most relevant vision concepts from complex inputs, based on task-specific instructions. To optimize this vision concept modeling process, we propose VCM, a self-supervised framework that leverages vision-language correlations across diverse instances. VCM is designed to learn meaningful vision concepts without the need for expensive concept-level annotations. At its core, it employs a forward-backward optimization algorithm that supports LVLMs to adjust concept granularity and spatial alignment dynamically. Experiments demonstrate that VCM remarkably reduces computational costs (e.g., achieving up to 85\% fewer FLOPs for LLaVA-1.5-7B), while maintaining strong performance across a series of vision-language tasks. The codebase is available at https://github.com/RainBowLuoCS/VCM.

In this work, we aim to incentivize the reasoning ability of Multimodal Large Language Models (MLLMs) via reinforcement learning (RL) and develop an effective approach that mitigates the sparse reward and advantage vanishing issues during RL. To this end, we propose Share-GRPO, a novel RL approach that tackle these issues by exploring and sharing diverse reasoning trajectories over expanded question space. Specifically, Share-GRPO first expands the question space for a given question via data transformation techniques, and then encourages MLLM to effectively explore diverse reasoning trajectories over the expanded question space and shares the discovered reasoning trajectories across the expanded questions during RL. In addition, Share-GRPO also shares reward information during advantage computation, which estimates solution advantages hierarchically across and within question variants, allowing more accurate estimation of relative advantages and improving the stability of policy training. Extensive evaluations over 6 widely-used reasoning benchmarks showcase the superior performance of our method. Code is available at https://github.com/HJYao00/R1-ShareVL.

Recent advancements in large reasoning models (LRMs) have introduced an intermediate "thinking" process prior to generating final answers, improving their reasoning capabilities on complex downstream tasks. However, the potential of LRMs as evaluators for machine translation (MT) quality remains underexplored. We provides the first systematic analysis of LRM-as-a-judge in MT evaluation. We identify key challenges, revealing LRMs require tailored evaluation materials, tend to "overthink" simpler instances and have issues with scoring mechanisms leading to overestimation. To address these, we propose to calibrate LRM thinking by training them on synthetic, human-like thinking trajectories. Our experiments on WMT24 Metrics benchmarks demonstrate that this approach largely reduces thinking budgets by ~35x while concurrently improving evaluation performance across different LRM scales from 7B to 32B (e.g., R1-Distill-Qwen-7B achieves a +8.7 correlation point improvement). These findings highlight the potential of efficiently calibrated LRMs to advance fine-grained automatic MT evaluation.

Conversational stance detection, which aims to identify stances within conversation threads, has become a research hotspot recently. As the number of dialogue turns increases and the conversation content becomes more complex, existing methods that simply incorporate conversational context are insufficient to effectively capture the nuanced information necessary for accurate stance detection. To address this issue, we introduce a Multi-agent Reasoning Graph Knowledge Distillation (MRGKD) framework, leveraging conversational reasoning among multiple Large Language Models (LLMs) into smaller language models. Specifically, we first construct a multi-agent reasoning graph to infer implicit logical relationships within the conversational history from the diverse perspectives of multiple LLMs. To fully leverage the in-context learning capabilities of LLMs, we design a reasoning knowledge editing mechanism that internalizes new information by aligning the output distribution of smaller language models with both the conversational history and the knowledge derived from the multi-agent reasoning graph. Additionally, we incorporate a contrastive loss to distinguish between correct and incorrect reasoning, alongside a stance detection loss, to fine-tune the smaller language models. This approach not only ensures the accurate acquisition of logical knowledge but also preserves the integrity of the conversational history. Experiments conducted on two public datasets demonstrate that our MRGKD significantly outperforms all baselines.

Abstract Detecting content generated by large language models (LLMs) is crucial for preventing misuse and building trustworthy AI systems. Although existing detection methods perform well, their robustness in out-of-distribution (OOD) scenarios is still lacking. In this paper, we hypothesize that, compared to features used by existing detection methods, the internal representations of LLMs contain more comprehensive and raw features that can more effectively capture and distinguish the statistical pattern differences between LLM-generated texts (LGT) and human-written texts (HWT). We validated this hypothesis across different LLMs and observed significant differences in neural activation patterns when processing these two types of texts. Based on this, we propose RepreGuard, an efficient statistics-based detection method. Specifically, we first employ a surrogate model to collect representation of LGT and HWT, and extract the distinct activation feature that can better identify LGT. We can classify the text by calculating the projection score of the text representations along this feature direction and comparing with a precomputed threshold. Experimental results show that RepreGuard outperforms all baselines with average 94.92% AUROC on both in-distribution and OOD scenarios, while also demonstrating robust resilience to various text sizes and mainstream attacks.1

Abstract Recent advances in large language models (LLMs) focus on aligning models with human values to minimize harmful content. However, existing methods often rely on a single type of feedback, such as preferences, annotated labels, or critiques, which can lead to overfitting and suboptimal performance. In this paper, we propose Diverse AIFeedback (DAIF), a novel approach that integrates three types of feedback—critique, refinement, and preference—tailored to tasks of varying uncertainty levels. Through an analysis of information gain, we show that critique feedback is most effective for low-uncertainty tasks, refinement feedback for medium-uncertainty tasks, and preference feedback for high-uncertainty tasks. Training with this diversified feedback reduces overfitting and improves alignment. Experimental results across three tasks—question answering, dialog generation, and text summarization–demonstrate that DAIF outperforms traditional methods relying on a single feedback type.1

Multimodal Emotion Recognition in Conversations (MERC) aims to identify the emotion expressed in each utterance within conversational videos. Current efforts are directed toward modeling speaker-sensitive context dependencies and multimodal fusion. However, they still struggle to handle utterances from unseen speakers, hampering the model's generalizability. To tackle this challenge, we propose a Speaker Generalization Framework for MERC. Specifically, we build a prototype graph to learn Speaker-based Utterance Representations (SUR), leveraging prototypes as the bridge between seen and unseen speakers. Speaker-aware Contrastive Learning (CL) is then applied to refine SUR, pulling utterances (or prototypes) from the same speaker together while pushing those from different speakers apart. Further, we introduce a prototypical graph CL to generalize SUR to unseen speakers, ensuring that the same speakers exhibit similar graph structures, while dissimilar ones differ. To further enhance model generalization, we introduce Uncertainty-based Generalization for Speakers, randomly sampling SUR statistics from the estimated Gaussian distribution and probabilistically replacing the original SUR. Experimental findings on two datasets highlight that our framework substantially improves the generalization of various MERC models, surpassing state-of-the-art methods.

Sexism has become a pressing issue, driven by the rapid-spreading influence of societal norms, media portrayals, and online platforms that perpetuate and amplify gender biases. Curbing sexism has emerged as a critical challenge globally. Being capable of recognizing sexist statements and behaviors is of particular importance since it is the first step in mind change. This survey provides an extensive overview of recent advancements in sexism detection. We present details of the various resources used in this field and methodologies applied to the task, covering different languages, modalities, models, and approaches. Moreover, we examine the specific challenges these models encounter in accurately identifying and classifying sexism. Additionally, we highlight areas that require further research and propose potential new directions for future exploration in the domain of sexism detection. Through this comprehensive exploration, we strive to contribute to the advancement of interdisciplinary research, fostering a collective effort to combat sexism in its multifaceted manifestations.

Relation extraction (RE) tends to struggle when the supervised training data is few and difficult to be collected. In this article, we elicit relational and factual knowledge from large pretrained language models (PLMs) for few-shot RE (FSRE) with prompting techniques. Concretely, we automatically generate a diverse set of natural language templates and modulate PLM's behavior through these prompts for FSRE. To mitigate the template bias which leads to unstableness of few-shot learning, we propose a simple yet effective template regularization network (TRN) to prevent deep networks from over-fitting uncertain templates and thus stabilize the FSRE models. TRN alleviates the template bias with three mechanisms: 1) an attention mechanism over mini-batch to weight each template; 2) a ranking regularization mechanism to regularize the attention weights and constrain the importance of uncertain templates; and 3) a template calibration module with two calibrating techniques to modify the uncertain templates in the lowest-ranked group. Experimental results on two benchmark datasets (i.e., FewRel and NYT) show that our model has robust superiority over strong competitors. For reproducibility, we will release our code and data upon the publication of this article.

Despite having achieved notable success for aspect-based sentiment analysis (ABSA), deep neural networks are susceptible to spurious correlations between input features and output labels, leading to poor robustness. In this paper, we propose a novel Counterfactual-Enhanced Information Bottleneck framework (called CEIB) to reduce spurious correlations for ABSA. CEIB extends the information bottleneck (IB) principle to a factual-counterfactual balancing setting by integrating augmented counterfactual data, with the goal of learning a robust ABSA model. Concretely, we first devise a multi-pattern prompting method, which utilizes the large language model (LLM) to generate high-quality counterfactual samples from the original samples. Then, we employ the information bottleneck principle and separate the mutual information into factual and counterfactual parts. In this way, we can learn effective and robust representations for the ABSA task by balancing the predictive information of these two parts. Extensive experiments on five benchmark ABSA datasets show that our CEIB approach achieves superior prediction performance and robustness over the state-of-the-art baselines. Code and data to reproduce the results in this paper is available at: https://github.com/shesshan/CEIB.

Multi-object tracking (MOT) is a challenging vision task that aims to detect individual objects within a single frame and associate them across multiple frames. Recent MOT approaches can be categorized into two-stage tracking-by-detection (TBD) methods and one-stage joint detection and tracking (JDT) methods. Despite the success of these approaches, they also suffer from common problems, such as harmful global or local inconsistency, poor trade-off between robustness and model complexity, and lack of flexibility in different scenes within the same video. In this paper we propose a simple but robust framework that formulates object detection and association jointly as a consistent denoising diffusion process from paired noise boxes to paired ground-truth boxes. This novel progressive denoising diffusion strategy substantially augments the tracker's effectiveness, enabling it to discriminate between various objects. During the training stage, paired object boxes diffuse from paired ground-truth boxes to random distribution, and the model learns detection and tracking simultaneously by reversing this noising process. In inference, the model refines a set of paired randomly generated boxes to the detection and tracking results in a flexible one-step or multi-step denoising diffusion process. Extensive experiments on three widely used MOT benchmarks, including MOT17, MOT20, and DanceTrack, demonstrate that our approach achieves competitive performance compared to the current state-of-the-art methods. Code is available at https://github.com/RainBowLuoCS/DiffusionTrack.

Longze Chen, Ziqiang Liu, Wanwei He, Yinhe Zheng, Hao Sun, Yunshui Li, Run Luo, Min Yang. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Large Language Models (LLMs) exhibit substantial capabilities yet encounter challenges, including hallucination, outdated knowledge, and untraceable reasoning processes.Retrievalaugmented generation (RAG) has emerged as a promising solution, integrating knowledge from external databases to mitigate these challenges.However, inappropriate retrieved passages can potentially hinder the LLMs' capacity to generate comprehensive and high-quality responses.Prior RAG studies on the robustness of retrieval noises often confine themselves to a limited set of noise types, deviating from realworld retrieval environments and limiting practical applicability.In this study, we initially investigate retrieval noises and categorize them into three distinct types, reflecting real-world environments.We analyze the impact of these various retrieval noises on the robustness of LLMs.Subsequently, we propose a novel RAG approach known as Retrieval-augmented Adaptive Adversarial Training (RAAT).RAAT leverages adaptive adversarial training to dynamically adjust the model's training process in response to retrieval noises.Concurrently, it employs multi-task learning to ensure the model's capacity to internally recognize noisy contexts.Extensive experiments demonstrate that the LLaMA-2 7B model trained using RAAT exhibits significant improvements in F1 and EM scores under diverse noise conditions.For reproducibility, we release our code and data at: https://github.com/calubkk/RAAT.

Yunshui Li, Binyuan Hui, Xiaobo Xia, Jiaxi Yang, Min Yang, Lei Zhang, Shuzheng Si, Ling-Hao Chen, Junhao Liu, Tongliang Liu, Fei Huang, Yongbin Li. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Recent advancements in Large Language Models (LLMs) have showcased their remarkable capabilities in text understanding and generation.However, even stronger LLMs are susceptible to acquiring erroneous or obsolete information from the training corpus.Direct secondary fine-tuning with data containing new knowledge may be ineffective in updating knowledge due to the conflict between old and new knowledge.In this paper, we propose a new paradigm for fine-tuning called F-Learning (Forgetting before Learning), which employs parametric arithmetic to facilitate the forgetting of old knowledge and learning of new knowledge.Experimental results on two publicly available datasets demonstrate that our proposed F-Learning can obviously improve the knowledge updating performance of both full fine-tuning and LoRA fine-tuning, simultaneously outperforming the existing baselines in most cases.Moreover, we have also discovered that forgetting old knowledge by subtracting the parameters of LoRA can yield a similar effect to subtracting the parameters of full fine-tuning, and occasionally even surpass it significantly.

Yang Sun, Muyi Wang, Jianzhu Bao, Bin Liang, Xiaoyan Zhao, Caihua Yang, Min Yang, Ruifeng Xu. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Jiaxi Yang, Binyuan Hui, Min Yang, Bailin Wang, Bowen Li, Binhua Li, Fei Huang, Yongbin Li. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

The capability gap between open-source and closed-source large language models (LLMs) remains challenging in text-to-SQL tasks.In this paper, we introduce a synthetic data approach that amalgamates strong data generated by larger, more potent models (strong models) with weak data produced by smaller, less wellaligned models (weak models).Our approach contributes to the improvement of domain generalization in text-to-SQL models and investigates the potential of weak data supervision through preference learning.Moreover, we utilize the synthetic data approach for instruction tuning on open-source LLMs, yielding SENSE, a specialized text-to-SQL model.The effectiveness of SENSE is substantiated by achieving state-of-the-art results on the SPIDER and BIRD benchmarks, thereby mitigating the performance disparity between open-source models and the methods derived from closed-source models.

Lei Zhang, Yunshui Li, Ziqiang Liu, Jiaxi Yang, Junhao Liu, Longze Chen, Run Luo, Min Yang. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Empathetic response generation is designed to comprehend the emotions of others and select the most appropriate strategies to assist them in resolving emotional challenges. Empathy can be categorized into cognitive empathy and affective empathy. The former pertains to the ability to understand and discern the emotional issues and situations of others, while the latter involves the capacity to provide comfort. To enhance one's empathetic abilities, it is essential to develop both these aspects. Therefore, we develop an innovative framework that combines retrieval augmentation and emotional support strategy integration. Our framework starts with the introduction of a comprehensive emotional palette for empathy. We then apply appraisal theory to decompose this palette and create a database of empathetic responses. This database serves as an external resource and enhances the LLM's empathy by integrating semantic retrieval mechanisms. Moreover, our framework places a strong emphasis on the proper articulation of response strategies. By incorporating emotional support strategies, we aim to enrich the model's capabilities in both cognitive and affective empathy, leading to a more nuanced and comprehensive empathetic response. Finally, we extract datasets ED and ET from the empathetic dialogue dataset EmpatheticDialogues and ExTES based on dialogue length. Experiments demonstrate that our framework can enhance the empathy ability of LLMs from both cognitive and affective empathy perspectives. Our code is released at https://github.com/CAS-SIAT-XinHai/APTNESS.

Traditional legal retrieval systems designed to retrieve legal documents, statutes, precedents, and other legal information are unable to give satisfactory answers due to lack of semantic understanding of specific questions. Large Language Models (LLMs) have achieved excellent results in a variety of natural language processing tasks, which inspired us that we train a LLM in the legal domain to help legal retrieval. However, in the Chinese legal domain, due to the complexity of legal questions and the rigour of legal articles, there is no legal large model with satisfactory practical application yet. In this paper, we present DeliLaw, a Chinese legal counselling system based on a large language model. DeliLaw integrates a legal retrieval module and a case retrieval module to overcome the model hallucination. Users can consult professional legal questions, search for legal articles and relevant judgement cases, etc. on the DeliLaw system in a dialogue mode. In addition, DeliLaw supports the use of English for counseling. we provide the address of the system: https://data.delilegal.com/lawQuestion.

In recent years, there has been a trend in the field of recommender systems towards multi-task modeling and multi-scenario modeling. The aim is to enhance the performance of various tasks and scenarios by jointly training on multiple tasks or scenarios to learn common patterns and features. Joint modeling of tasks and scenarios has also received widespread attention recently. However, despite the rich proposals of methods for Multi-Task Learning (MTL), Multi-Scenario Learning (MSL), and Multi-Task-Multi-Scenario Learning (MTMSL) in recent years, there still lacks a comprehensive benchmark to evaluate these methods. Previous studies often employed different datasets, data processing techniques, data partitioning strategies, and hyperparameter settings, making replication of existing research and fair comparison of experimental results challenging. To address this challenge, we introduce MMLRec, the first unified comprehensive benchmark for evaluating MTL, MSL and MTMSL, featuring consistent dataset processing and identical parameter settings. This benchmark implements a range of MTL, MSL, and MTMSL algorithms, and evaluates them on multiple commonly used recommender systems datasets. Through fair comparative experiments, we find that some structurally simplistic recommendation algorithms are underestimated, as they can achieve comparable results to more complex algorithms while maintaining lower complexity. Furthermore, our experimental analysis indicates that more complex methods exhibit better robustness when there are significant differences between tasks or scenarios. By providing a unified framework (MMLRec), our goal is to promote rapid evaluation and inspire innovative research in this continuously evolving field. We hope that our open-source benchmark can facilitate swift, equitable evaluations, while also fostering further breakthrough research in the domains of MTL, MSL, and MTMSL.

Reinforcement learning from human feedback (RLHF) is a crucial technique in aligning large language models (LLMs) with human preferences, ensuring these LLMs behave in beneficial and comprehensible ways to users. However, a longstanding challenge in human alignment techniques based on reinforcement learning lies in their inherent complexity and difficulty in training. To address this challenge, we present a simple yet effective Contrastive Learning Framework for Human Alignment (CLHA) to align LLMs with human preferences directly. CLHA employs a novel rescoring strategy to evaluate the noise within the data by considering its inherent quality and dynamically adjusting the training process. Simultaneously, CLHA utilizes pairwise contrastive loss and adaptive supervised fine-tuning loss to adaptively modify the likelihood of generating responses, ensuring enhanced alignment with human preferences. Using advanced methods, CLHA surpasses other algorithms, showcasing superior performance in terms of reward model scores, automatic evaluations, and human assessments on the widely used “Helpful and Harmless” dataset.

Task-oriented dialogue (TOD) systems facilitate users in executing various activities via multi-turn dialogues, but Large Language Models (LLMs) often struggle to comprehend these intricate contexts. In this study, we propose a novel “Self-Explanation” prompting strategy to enhance the comprehension abilities of LLMs in multi-turn dialogues. This task-agnostic approach requires the model to analyze each dialogue utterance before task execution, thereby improving performance across various dialogue-centric tasks. Experimental results from six benchmark datasets confirm that our method consistently outperforms other zero-shot prompts and matches or exceeds the efficacy of few-shot prompts, demonstrating its potential as a powerful tool in enhancing LLMs’ comprehension in complex dialogue tasks.

In this paper, we introduce an innovative pre-training framework TP-Link, which aims to improve context-dependent Text-to-SQL Parsing by leveraging Linking information. This enhancement is achieved through better representation of both natural language utterances and the database schema, ultimately facilitating more effective text-to-SQL conversations. We present two novel pre-training objectives: (i) utterance linking prediction (ULP) task that models intricate syntactic relationships among natural language utterances in context-dependent text-to-SQL scenarios, and (ii) schema linking prediction (SLP) task that focuses on capturing fine-grained schema linking relationships between the utterances and the database schema. Extensive experiments demonstrate that our proposed TP-Link achieves state-of-the-art performance on two leading downstream benchmarks (i.e., SParC and CoSQL).

Large language models (LLMs) have demonstrated impressive performance in various natural language processing (NLP) tasks. However, there is limited understanding of how well LLMs perform in specific domains (e.g, the intellectual property (IP) domain). In this paper, we contribute a new benchmark, the first Multilingual-oriented quiZ on Intellectual Property (MoZIP), for the evaluation of LLMs in the IP domain. The MoZIP benchmark includes three challenging tasks: IP multiple-choice quiz (IPQuiz), IP question answering (IPQA), and patent matching (PatentMatch). In addition, we also develop a new IP-oriented multilingual large language model (called MoZi), which is a BLOOMZ-based model that has been supervised fine-tuned with multilingual IP-related text data. We evaluate our proposed MoZi model and four well-known LLMs (i.e., BLOOMZ, BELLE, ChatGLM and ChatGPT) on the MoZIP benchmark. Experimental results demonstrate that MoZi outperforms BLOOMZ, BELLE and ChatGLM by a noticeable margin, while it had lower scores compared with ChatGPT. Notably, the performance of current LLMs on the MoZIP benchmark has much room for improvement, and even the most powerful ChatGPT does not reach the passing level. Our source code, data, and models are available at https://github.com/AI-for-Science/MoZi.

Among the various pre-trained neural language models that are popular today, dropout is already an indispensable regularization technique. To solve the inconsistency between training and inference caused by the randomness of dropout, some studies use consistency training to regularize dropout at the output layer. In this paper, we propose a novel Layer-wise Regularized Dropout (LR-Drop), which is specially designed for Transformer-based Language models. Specifically, LR-Drop layer-wise regularizes each Transformer layer using the consistency training strategy. Each training sample passes through the two siamese sub-models sampled by dropout, and then LR-Drop forces the hidden states, multi-head attention matrices, and output distribution of the two siamese sub-models to be consistent. The proposed LR-Drop can be regarded as a “self-distillation” framework, in which each sub-model generated by dropout is the other’s “teacher” model and “student” model. Through extensive experiments on 8 natural language understanding datasets, 6 neural machine translation datasets, and 1 abstractive summarization dataset (a total of 15 datasets), we show that LR-Drop achieves superior performances, including state-of-the-art results.

Previous stance detection studies typically concentrate on evaluating stances within individual instances, thereby exhibiting limitations in effectively modeling multi-party discussions concerning the same specific topic, as naturally transpire in authentic social media interactions. This constraint arises primarily due to the scarcity of datasets that authentically replicate real social media contexts, hindering the research progress of conversational stance detection. In this paper, we introduce a new multi-turn conversation stance detection dataset (called MT-CSD), which encompasses multiple targets for conversational stance detection. To derive stances from this challenging dataset, we propose a global-local attention network (GLAN) to address both long and short-range dependencies inherent in conversational data. Notably, even state-of-the-art stance detection methods, exemplified by GLAN, exhibit an accuracy of only 50.47%, highlighting the persistent challenges in conversational stance detection. Furthermore, our MT-CSD dataset serves as a valuable resource to catalyze advancements in cross-domain stance detection, where a classifier is adapted from a different yet related target. We believe that MT-CSD will contribute to advancing real-world applications of stance detection research. Our source code, data, and models are available at https://github.com/nfq729/MT-CSD.

In recent years, instruction fine-tuning (IFT) on large language models (LLMs) has garnered considerable attention to enhance model performance on unseen tasks.Attempts have been made on automatic construction and effective selection for IFT data.However, we posit that previous methods have not fully harnessed the potential of LLMs for enhancing data quality.The responses within IFT data could be further enhanced by leveraging the capabilities of LLMs themselves.In this paper, we propose COEVOL, an LLM-based multiagent cooperation framework for the improvement of responses for instructions.To effectively refine the responses, we develop an iterative framework following a debate-adviseedit-judge paradigm.A two-stage multi-agent debate strategy is further devised to ensure the diversity and reliability of editing suggestions within the framework.Empirically, models equipped with COEVOL outperform competitive baselines evaluated by MT-Bench and Al-pacaEval, demonstrating its effectiveness in enhancing instruction-following capabilities for LLMs. 1

Minzheng Wang, Longze Chen, Fu Cheng, Shengyi Liao, Xinghua Zhang, Bingli Wu, Haiyang Yu, Nan Xu, Lei Zhang, Run Luo, Yunshui Li, Min Yang, Fei Huang, Yongbin Li. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing. 2024.

Recently, substantial advancements in pre-trained visionlanguage models have greatly enhanced the capabilities of multi-modal dialog systems. These models have demonstrated significant improvements by fine-tuning on downstream tasks. However, the existing pre-trained models primarily focus on effectively capturing the alignment between vision and language modalities, often ignoring the intricate nature of dialog context. In this paper, we propose a parameter-efficient prompt-tuning method named DialCLIP for multi-modal dialog retrieval. Specifically, our approach introduces a multi-modal context prompt generator to learn context features which are subsequently distilled into prompts within the pre-trained vision-language model CLIP. Besides, we introduce domain prompt to mitigate the disc repancy from the downstream dialog data. To facilitate various types of retrieval, we also design multiple experts to learn mappings from CLIP outputs to multi-modal representation space, with each expert being responsible to one specific retrieval type. Extensive experiments show that DialCLIP achieves stateof-the-art performance on two widely recognized benchmark datasets (i.e., PhotoChat and MMDialog) by tuning a mere 0.04% of the total parameters. These results highlight the efficacy and efficiency of our proposed approach, underscoring its potential to advance the field of multi-modal dialog retrieval.

The approach of Reinforcement Learning from Human Feedback (RLHF) is widely used for enhancing pre-trained Language Models (LM), enabling them to better align with human preferences. Existing RLHF-based LMs however require complete retraining whenever new queries or feedback are introduced, as human preferences may differ across different domains or topics. LM retraining is often impracticable in most real-world scenarios, due to the substantial time and computational costs involved, as well as data privacy concerns. To address this limitation, we propose Continual Proximal Policy Optimization (CPPO), a novel method that is able to continually align LM with dynamic human preferences. Specifically, CPPO adopts a weighting strategy to decide which samples should be utilized for enhancing policy learning and which should be used for solidifying past experiences. This seeks a good trade-off between policy learning and knowledge retention. Our experimental results show that CPPO outperforms strong Continuous learning (CL) baselines when it comes to consistently aligning with human preferences. Furthermore, compared to PPO, CPPO offers more efficient and stable learning in non-continual scenarios.

Image-text retrieval stands as a pivotal task within information retrieval, gaining increasing importance with the rapid advancements in Visual-Language Pretraining models. However, current benchmarks for evaluating these models face limitations, exemplified by instances such as BLIP2 achieving near-perfect performance on existing benchmarks. In response, this paper advocates for a more robust evaluation benchmark for image-text retrieval, one that embraces several essential characteristics. Firstly, a comprehensive benchmark should cover a diverse range of tasks in both perception and cognition-based retrieval. Recognizing this need, we introduce ReCoS, a novel benchmark specifically designed for cross-modal image-text retrieval in complex real-life scenarios. Unlike existing benchmarks, ReCoS encompasses 12 retrieval tasks, with a particular focus on three cognition-based tasks, providing a more holistic assessment of model capabilities. To ensure the novelty of the benchmark, we emphasize the use of original data sources, steering clear of reliance on existing publicly available datasets to minimize the risk of data leakage. Additionally, to strike a balance between the complexity of the real world and benchmark usability, ReCoS includes text descriptions that are neither overly detailed, making retrieval overly simplistic, nor under-detailed to the point where retrieval becomes impossible. Our evaluation results shed light on the challenges faced by existing methods, especially in cognition-based retrieval tasks within ReCoS. This underscores the necessity for innovative approaches in addressing the complexities of image-text retrieval in real-world scenarios. Our code and benchmark datasets are available for further research and development in this field https://github.com/Bruce-XJChen/ReCos

The rapid advancements in the development of multimodal large language models (MLLMs) have consistently led to new breakthroughs on various benchmarks. In response, numerous challenging and comprehensive benchmarks have been proposed to more accurately assess the capabilities of MLLMs. However, there is a dearth of exploration of the higher-order perceptual capabilities of MLLMs. To fill this gap, we propose the Image Implication understanding Benchmark, II-Bench, which aims to evaluate the model's higher-order perception of images. Through extensive experiments on II-Bench across multiple MLLMs, we have made significant findings. Initially, a substantial gap is observed between the performance of MLLMs and humans on II-Bench. The pinnacle accuracy of MLLMs attains 74.8%, whereas human accuracy averages 90%, peaking at an impressive 98%. Subsequently, MLLMs perform worse on abstract and complex images, suggesting limitations in their ability to understand high-level semantics and capture image details. Finally, it is observed that most models exhibit enhanced accuracy when image sentiment polarity hints are incorporated into the prompts. This observation underscores a notable deficiency in their inherent understanding of image sentiment. We believe that II-Bench will inspire the community to develop the next generation of MLLMs, advancing the journey towards expert artificial general intelligence (AGI). II-Bench is publicly available at https://huggingface.co/datasets/m-a-p/II-Bench.

Abstract To address the issue of low efficiency caused by the repeated use of quantum attack resistant static identity authentication methods in a satellite integrated smart grid, this paper proposes a quantum attack resistant continuous identity authentication protocol. First, in the initial authentication stage, in order to reduce computational complexity, the key encryption mechanism in the CRYSTALS‐Kyber algorithm was improved and combined with the NTRU message recovery digital signature scheme to construct a lattice based explicit AKE (Kyber NTRU. AKE), which achieved mutual authentication and negotiated shared tokens. Second, in the continuous authentication stage, incorporating quantum attack resistant tokens into the current algorithm to improve authentication efficiency. The formal analysis results indicate that compared to the weakly forward secure Kyber.AKE in the CRYSTALS‐Kyber algorithm, Kyber‐NTRU.AKE achieves complete forward secrecy, while the non‐formal analysis results demonstrate the security of the continuous authentication phase. Through theoretical analysis and efficiency comparison with Cyber.AKE, the analysis shows that the Cyber‐NTRU.AKE has higher computational and communication efficiency than Cyber.AKE.

Fine-grained meme understanding aims to explore and comprehend the meanings of memes from multiple perspectives by performing various tasks, such as sentiment analysis, intention detection, and offensiveness detection. Existing approaches primarily focus on simple multi-modality fusion and individual task analysis. However, there remain several limitations that need to be addressed: (1) the neglect of incongruous features within and across modalities, and (2) the lack of consideration for correlations among different tasks. To this end, we leverage metaphorical information as text modality and propose a Metaphor-aware Multi-modal Multi-task Framework (M3F) for fine-grained meme understanding. Specifically, we create inter-modality attention enlightened by the Transformer to capture inter-modality interaction between text and image. Moreover, intra-modality attention is applied to model the contradiction between the text and metaphorical information. To learn the implicit interaction among different tasks, we introduce a multi-interactive decoder that exploits gating networks to establish the relationship between various subtasks. Experimental results on the MET-Meme dataset show that the proposed framework outperforms the state-of-the-art baselines in fine-grained meme understanding.

Abstract Data-to-text (D2T) generation aims to transform structured data into natural language text. Data-to-text pre-training has proved to be powerful in enhancing D2T generation and yields impressive performance. However, previous pre-training methods either oversimplified structured data into a sequence without considering input structures or designed training objectives tailored for a specific data structure (e.g., table or knowledge graph). In this paper, we unify different types of structured data (i.e., table, key-value data, knowledge graph) into the graph format and cast different D2T generation tasks as graph-to-text generation. To effectively exploit the structural information of the input graph, we propose a structure-enhanced pre-training method for D2T generation by designing a structure-enhanced Transformer. Concretely, we devise a position matrix for the Transformer, encoding relative positional information of connected nodes in the input graph. In addition, we propose a new attention matrix to incorporate graph structures into the original Transformer by taking the available explicit connectivity structure into account. Extensive experiments on six benchmark datasets show the effectiveness of our model. Our source codes are available at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/unid2t.

Multimodal aspect-based sentiment analysis (MABSA) aims to determine the sentiment polarity of each aspect mentioned in the text based on multimodal content. Various approaches have been proposed to model multimodal sentiment features for each aspect via modal interactions. However, most existing approaches have two shortcomings: (1) The representation gap between textual and visual modalities may increase the risk of misalignment in modal interactions; (2) In some examples where the image is not related to the text, the visual information may not enrich the textual modality when learning aspect-based sentiment features. In such cases, blindly leveraging information from visual modal may introduce noises in reasoning the aspect-based sentiment expressions. To tackle the shortcomings mentioned above, we propose an end-to-end MABSA framework with image conversion and noise filtration. Specifically, to bridge the representation gap in different modalities, we attempt to translate images into the input space of a pre-trained language model (PLM). To this end, we develop an image-to-text conversion module that can convert an image to an implicit sequence of token embedding. Moreover, an aspect-oriented filtration module is devised to alleviate the noise in the implicit token embeddings, which consists of two attention operations. The former aims to create an enhanced aspect embedding as a query, and the latter seeks to use this query to retrieve relevant auxiliary information from the implicit token embeddings to supplement the textual content. After filtering the noise, we leverage a PLM to encode the text, aspect, and image prompt derived from filtered implicit token embeddings as sentiment features to perform aspect-based sentiment prediction. Experimental results on two MABSA datasets show that our framework achieves state-of-the-art performance. Furthermore, extensive experimental analysis demonstrates the proposed framework has superior robustness and efficiency.

Adversarial training has been proven to be a powerful regularization technique to improve language models. In this work, we propose a novel randomdropped weightattackadversarial training method (DropAttack) for natural language understanding. Our DropAttack improves the generalization of models by minimizing the internal adversarial risk caused by a multitude of attack combinations. Specifically, DropAttack enhances the adversarial attack space by intentionally adding worst-case adversarial perturbations to the weight parameters and randomly dropping the specific proportion of attack perturbations. To extensively validate the effectiveness of DropAttack,12public English natural language understanding datasets were used. Experiments on the GLUE benchmark show that when DropAttack is applied only to the finetuning stage, it is able to improve the overall test scores of the BERT-base pre-trained model from 78.3 to 79.7 and RoBERTa-large pre-trained model from 88.1 to 88.8. Further, DropAttack also significantly improves models trained from scratch. Theoretical analysis reveals that DropAttack performs potential gradient regularization on the input and weight parameters of the model. Moreover, visualization experiments show that DropAttack can push the minimum risk of the neural network to a lower and flatter loss landscape.

The task of entity alignment between knowledge graphs (KGs) aims to identify every pair of entities from two different KGs that represent the same entity. Many machine learning-based methods have been proposed for this task. However, to our best knowledge, existing methods all requiremanually craftedseed alignments, which are expensive to obtain. In this paper, we propose the first fully automatic alignment method named AutoAlign, which does not require any manually crafted seed alignments. Specifically, for predicate embeddings, AutoAlign constructs a predicate-proximity-graph with the help of large language models to automatically capture the similarity between predicates across two KGs. For entity embeddings, AutoAlign first computes the entity embeddings of each KG independently using TransE, and then shifts the two KGs' entity embeddings into the same vector space by computing the similarity between entities based on their attributes. Thus, both predicate alignment and entity alignment can be done without manually crafted seed alignments. AutoAlign is not only fully automatic, but also highly effective. Experiments using real-world KGs show that AutoAlign improves the performance of entity alignment significantly compared to state-of-the-art methods. Our source code is available at ruizhang-ai/AutoAlign.

The goal of semi-supervised text classification (SSTC) is to train a model by exploring both a small number of labeled data and a large number of unlabeled data, such that the learned semi-supervised classifier performs better than the supervised classifier trained on solely the labeled samples. Pseudo-labeling is one of the most widely used SSTC techniques, which trains a teacher classifier with a small number of labeled examples to predict pseudo labels for the unlabeled data. The generated pseudo-labeled examples are then utilized to train a student classifier, such that the learned student classifier can outperform the teacher classifier. Nevertheless, the predicted pseudo labels may be inaccurate, making the performance of the student classifier degraded. The student classifier may perform even worse than the teacher classifier. To alleviate this issue, in this paper, we introduce a dual meta-learning ( DML ) technique for semi-supervised text classification, which improves the teacher and student classifiers simultaneously in an iterative manner. Specifically, we propose a meta-noise correction method to improve the student classifier by proposing a Noise Transition Matrix (NTM) with meta-learning to rectify the noisy pseudo labels. In addition, we devise a meta pseudo supervision method to improve the teacher classifier. Concretely, we exploit the feedback performance from the student classifier to further guide the teacher classifier to produce more accurate pseudo labels for the unlabeled data. In this way, both teacher and student classifiers can co-evolve in the iterative training process. Extensive experiments on four benchmark datasets highlight the effectiveness of our DML method against existing state-of-the-art methods for semi-supervised text classification. We release our code and data of this paper publicly at https://github.com/GRIT621/DML.

Open intent classification, which aims to correctly classify the known intents into their corresponding classes while identifying the new unknown (open) intents, is an essential but challenging task in dialogue systems. In this paper, we introduce novel K-center contrastive learning and adjustable decision boundary learning (CLAB) to improve the effectiveness of open intent classification. First, we pre-train a feature encoder on the labeled training instances, which transfers knowledge from known intents to unknown intents. Specifically, we devise a K-center contrastive learning algorithm to learn discriminative and balanced intent features, improving the generalization of the model for recognizing open intents. Second, we devise an adjustable decision boundary learning method with expanding and shrinking (ADBES) to determine the suitable decision conditions. Concretely, we learn a decision boundary for each known intent class, which consists of a decision center and the radius of the decision boundary. We then expand the radius of the decision boundary to accommodate more in-class instances if the out-of-class instances are far from the decision boundary; otherwise, we shrink the radius of the decision boundary. Extensive experiments on three benchmark datasets clearly demonstrate the effectiveness of our method for open intent classification.For reproducibility, we submit the code at: https://github.com/lxk00/CLAP

The ability to continually learn over time by grasping new knowledge and remembering previously learned experiences is essential for developing an online task-oriented dialogue system (TDS). In this paper, we work on the class incremental learning scenario where the TDS is evaluated without specifying the dialogue domain. We employ contrastive distillation on the intermediate representations of dialogues to learn transferable representations that suffer less from catastrophic forgetting. Besides, we provide a dynamic update mechanism to explicitly preserve the learned experiences by only updating the parameters related to the new task while keeping other parameters fixed. Extensive experiments demonstrate that our method significantly outperforms the strong baselines.

In real-world scenarios, it is crucial to build a lifelong taskoriented dialogue system (TDS) that continually adapts to new knowledge without forgetting previously acquired experiences. Existing approaches mainly focus on mitigating the catastrophic forgetting in lifelong TDS. However, the transfer ability to generalize the accumulated old knowledge to new tasks is underexplored. In this paper, we propose a two-stage lifelong task-oriented dialogue generation method to mitigate catastrophic forgetting and encourage knowledge transfer simultaneously, inspired by the learning process. In the first stage, we learn task-specific masks which adaptively preserve the knowledge of each visited task so as to mitigate catastrophic forgetting. In this stage, we are expected to learn the task-specific knowledge which is tailored for each task. In the second stage, we bring the knowledge from the encountered tasks together and understand thoroughly. To this end, we devise a balanced meta learning strategy for both forward and backward knowledge transfer in the lifelong learning process. In particular, we perform meta-update with a meta-test set sampled from the current training data for forward knowledge transfer. In addition, we employ an uncertainty-based sampling strategy to select and store representative dialogue samples into episodic memory and perform meta-update with a meta-test set sampled from the memory for backward knowledge transfer. With extensive experiments on 29 tasks, we show that MetaLTDS outperforms the strong baselines in terms of both effectiveness and efficiency. For reproducibility, we submit our code at: https: //github.com/travis-xu/MetaLTDS.

Script event prediction aims to predict the subsequent event given the context. This requires the capability to infer the correlations between events. Recent works have attempted to improve event correlation reasoning by using pretrained language models and incorporating external knowledge (e.g., discourse relations). Though promising results have been achieved, some challenges still remain. First, the pretrained language models adopted by current works ignore event-level knowledge, resulting in an inability to capture the correlations between events well. Second, modeling correlations between events with discourse relations is limited because it can only capture explicit correlations between events with discourse markers, and cannot capture many implicit correlations. To this end, we propose a novel generative approach for this task, in which a pretrained language model is fine-tuned with an event-centric pretraining objective and predicts the next event within a generative paradigm. Specifically, we first introduce a novel event-level blank infilling strategy as the learning objective to inject event-level knowledge into the pretrained language model, and then design a likelihood-based contrastive loss for fine-tuning the generative model. Instead of using an additional prediction layer, we perform prediction by using sequence likelihoods generated by the generative model. Our approach models correlations between events in a soft way without any external knowledge. The likelihood-based prediction eliminates the need to use additional networks to make predictions and is somewhat interpretable since it scores each word in the event. Experimental results on the multi-choice narrative cloze (MCNC) task demonstrate that our approach achieves better results than other state-of-the-art baselines. Our code will be available at https://github.com/zhufq00/mcnc.

Perceiving multi-modal information and fulfilling dialogues with humans is a long-term goal of artificial intelligence.Pre-training is commonly regarded as an effective approach for multi-modal dialogue.However, due to the limited availability of multi-modal dialogue data, there is still scarce research on multi-modal dialogue pre-training.Yet another intriguing challenge emerges from the encompassing nature of multi-modal dialogue, which involves various modalities and tasks.Moreover, new forms of tasks may arise at unpredictable points in the future.Hence, it is essential for designed multi-modal dialogue models to possess sufficient flexibility to adapt to such scenarios.This paper proposes PaCE, a unified, structured, compositional multi-modal dialogue pretraining framework.It utilizes a combination of several fundamental experts to accommodate multiple dialogue-related tasks and can be pre-trained using limited dialogue and extensive non-dialogue multi-modal data.Furthermore, we propose a progressive training method where old experts from the past can assist new experts, facilitating the expansion of their capabilities.Experimental results demonstrate that PaCE achieves state-of-the-art results on eight multi-modal dialog benchmarks.

Tianshu Yu, Haoyu Gao, Ting-En Lin, Min Yang, Yuchuan Wu, Wentao Ma, Chao Wang, Fei Huang, Yongbin Li. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Identifying users' stances regarding specific targets/topics is a significant route to learning public opinion from social media platforms.Most existing studies of stance detection strive to learn stance information about specific targets from the context, in order to determine the user's stance on the target.However, in real-world scenarios, we usually have a certain understanding of a target when we express our stance on it.In this paper, we investigate stance detection from a novel perspective, where the background knowledge of the targets is taken into account for better stance detection.To be specific, we categorize background knowledge into two categories: episodic knowledge and discourse knowledge, and propose a novel Knowledge-Augmented Stance Detection (KASD) framework.For episodic knowledge, we devise a heuristic retrieval algorithm based on the topic to retrieve the Wikipedia documents relevant to the sample.Further, we construct a prompt for ChatGPT to filter the Wikipedia documents to derive episodic knowledge.For discourse knowledge, we construct a prompt for ChatGPT to paraphrase the hashtags, references, etc., in the sample, thereby injecting discourse knowledge into the sample.Experimental results on four benchmark datasets demonstrate that our KASD achieves state-of-the-art performance in in-target and zero-shot stance detection.

Unintended dataset biases typically exist in existing Emotion Recognition in Conversations (ERC) datasets, including label bias, where models favor the majority class due to imbalanced training data, as well as the speaker and neutral word bias, where models make unfair predictions because of excessive correlations between specific neutral words or speakers and classes.However, previous studies in ERC generally focus on capturing context-sensitive and speaker-sensitive dependencies, ignoring the unintended dataset biases of data, which hampers the generalization and fairness in ERC.To address this issue, we propose a Training-Free Debiasing framework (TFD) that operates during prediction without additional training.To ensure compatibility with various ERC models, it does not balance data or modify the model structure.Instead, TFD extracts biases from the model by generating counterfactual utterances and contexts and mitigates them using simple yet empirically robust element-wise subtraction operations.Extensive experiments on three public datasets demonstrate that TFD effectively improves generalization ability and fairness across different ERC models 1 .

Multi-source domain adaptation (MSDA) aims to leverage the knowledge in multiple source domains to assist the prediction in a target domain, where the source and target domains have different data distributions. This paper presents a MSDA model to investigate both domain discrepancy and domain relevance, whose interactions are also exploited to gradually refine the learning performance. Particularly, the proposed model contains two components, i.e., feature spaces learning and transferred weights learning. The former one minimizes the domain discrepancy and the latter one evaluates the domain relevance. Experimental results on several real-world datasets demonstrate the effectiveness of the proposed model.

Action recognition research has gained significant attention with two dominant unimodal approaches: skeleton-based and RGB video-based. While the former is known for its robustness in complex backgrounds, the latter provides rich environmental information useful for context-based analysis. However, the fusion of these two modalities remains an open challenge. In this paper, we propose a Spatial Transformer & Selective Temporal encoder (ST&ST) for skeleton-based action recognition by constructing two modules: Reranking-Enhanced Dynamic Mask Transformer (RE-DMT) and Selective Kernel Temporal Convolution (SK-TC). The RE-DMT captures global spatial features, while the dynamic mask strategy and reranking strategy reduce redundancy. The SK-TC captures both long-term and short-term temporal features and enables adaptive fusion. Furthermore, in two phases, we propose a Homogeneous-Heterogeneous Multimodal Network (HHMNet) for multi-modal action recognition. In the first phase, contrastive learning is employed to achieve implicit semantic fusion within the four homogeneous skeletal modalities (joint, bone, etc.). In the second phase, the fusion of heterogeneous modalities (skeleton & RGB video) is carried out at three levels: model, feature, and decision. At the model level, the powerful skeleton-based model from the previous phase provides explicit attention guidance to the RGB video-based model. At the feature level, multi-part contrastive learning enables semantic distillation between heterogeneous modalities. At the decision level, ensemble learning combines outputs for final action recognition. We evaluate our proposed ST&ST guided HHMNet on NTU RGB+D 60 & 120 and NW-UCLA datasets and demonstrate that it achieves state-of-the-art performance in both skeleton-based and multi-modal action recognition tasks.

Intent detection plays an essential role in dialogue systems. This paper takes the lead to study open compound domain adaptation (OCDA) for intent detection, which brings the advantage of improved generalization to unseen domains. OCDA for intent detection is indeed a more realistic domain adaptation setting, which learns an intent classifier from labeled source domains and adapts it to unlabeled compound target domains containing different intent classes with the source domains. At inference time, we test the intent classifier in open domains that contain previously unseen intent classes. To this end, we propose an Adversarial Meta Prompt Tuning method (called AMPT) for open compound domain adaptive intent detection. Concretely, we propose a meta prompt tuning method, which utilizes language prompts to elicit rich knowledge from large-scale pre-trained language models (PLMs) and automatically finds better prompt initialization that facilitates fast adaptation via meta learning. Furthermore, we leverage a domain adversarial training technique to acquire domain-invariant representations of diverse domains. By taking advantage of the collaborative effect of meta learning, prompt tuning, and adversarial training, we can learn an intent classifier that can effectively generalize to unseen open domains. Experimental results on two benchmark datasets (i.e., HWU64 and CLINC) show that our model can learn substantially better-generalized representations for unseen domains compared with strong competitors.

Dialogue Topic Segmentation (DTS) plays an essential role in a variety of dialogue modeling tasks. Previous DTS methods either focus on semantic similarity or dialogue coherence to assess topic similarity for unsupervised dialogue segmentation. However, the topic similarity cannot be fully identified via semantic similarity or dialogue coherence. In addition, the unlabeled dialogue data, which contains useful clues of utterance relationships, remains underexploited. In this paper, we propose a novel unsupervised DTS framework, which learns topic-aware utterance representations from unlabeled dialogue data through neighboring utterance matching and pseudo-segmentation. Extensive experiments on two benchmark datasets (i.e., DialSeg711 and Doc2Dial) demonstrate that our method significantly outperforms the strong baseline methods. For reproducibility, we provide our code and data at: https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/dial-start.

Table-based reasoning has shown remarkable progress in a wide range of table-based tasks. It is a challenging task, which requires reasoning over both free-form natural language (NL) questions and (semi-)structured tabular data. However, previous table-based reasoning solutions usually suffer from significant performance degradation on ''huge'' evidence (tables). In addition, most existing methods struggle to reason over complex questions since the essential information is scattered in different places. To alleviate the above challenges, we exploit large language models (LLMs) as decomposers for effective table-based reasoning, which (i) decompose huge evidence (a huge table) into sub-evidence (a small table) to mitigate the interference of useless information for table reasoning, and (ii) decompose a complex question into simpler sub-questions for text reasoning. First, we use a powerful LLM to decompose the evidence involved in the current question into the sub-evidence that retains the relevant information and excludes the remaining irrelevant information from the ''huge'' evidence. Second, we propose a novel ''parsing-execution-filling'' strategy to decompose a complex question into simper step-by-step sub-questions by generating intermediate SQL queries as a bridge to produce numerical and logical sub-questions with a powerful LLM. Finally, we leverage the decomposed sub-evidence and sub-questions to get the final answer with a few in-context prompting examples. Extensive experiments on three benchmark datasets (TabFact, WikiTableQuestion, and FetaQA) demonstrate that our method achieves significantly better results than competitive baselines for table-based reasoning. Notably, our method outperforms human performance for the first time on the TabFact dataset. In addition to impressive overall performance, our method also has the advantage of interpretability, where the returned results are to some extent tractable with the generated sub-evidence and sub-questions. For reproducibility, we release our source code and data at: https://github.com/AlibabaResearch/DAMO-ConvAI.

Deep neural models have become mainstream in relation extraction (RE), yielding state-of-the-art performance. However, most existing neural models are prone to spurious correlations between input features and prediction labels, making the models suffer from low robustness and generalization.In this paper, we propose a spurious correlation reduction method for RE via feature decomposition and semantic augmentation (denoted as FDSA). First, we decompose the original sentence representation into class-related features and context-related features. To obtain better context-related features, we devise a contrastive learning method to pull together the context-related features of the anchor sentence and its augmented sentences, and push away the context-related features of different anchor sentences. In addition, we propose gradient-based semantic augmentation on context-related features in order to improve the robustness of the RE model. Experiments on four datasets show that our model outperforms the strong competitors.

The state-of-the-art approaches to targeted aspect-based sentiment analysis (TABSA) are mostly built on deep neural networks with attention mechanisms. One problem is that embeddings of targets and aspects are either pre-trained from large external corpora or randomly initialized. We argue that affective commonsense knowledge and words indicative of sentiment could be used to learn better target and aspect embeddings. We therefore propose an embedding refinement framework calledRAEC(RefiningAffectiveEmbedding fromContext), in which sentiment concepts extracted from affective commonsense knowledge and word relative location information are incorporated to derive context-affective embeddings. Furthermore, a sparse coefficient vector is exploited in refining the embeddings of targets and aspects separately. In this way, embeddings of targets and aspects can capture the highly relevant affective words. Experimental results on two benchmark datasets show that our framework can be easily integrated with existing embedding-based TABSA models and achieves state-of-the-art results compared to models relying on pre-trained word embeddings or built on other embedding refinement methods.

Making accurate recommendations for cold-start users has been a longstanding and critical challenge for recommender systems (RS). Cross-domain recommendations (CDR) offer a solution to tackle such a cold-start problem when there is no sufficient data for the users who have rarely used the system. An effective approach in CDR is to leverage the knowledge (e.g., user representations) learned from a related but different domain and transfer it to the target domain. Fine-tuning works as an effective transfer learning technique for this objective, which adapts the parameters of a pre-trained model from the source domain to the target domain. However, current methods are mainly based on the global fine-tuning strategy: the decision of which layers of the pre-trained model to freeze or fine-tune is taken for all users in the target domain. In this paper, we argue that users in RS are personalized and should have their own fine-tuning policies for better preference transfer learning. As such, we propose a novel User-specific Adaptive Fine-tuning method (UAF), selecting which layers of the pre-trained network to fine-tune, on a per-user basis. Specifically, we devise a policy network with three alternative strategies to automatically decide which layers to be fine-tuned and which layers to have their parameters frozen for each user. Extensive experiments show that the proposed UAF exhibits significantly better and more robust performance for user cold-start recommendation.

Recently, explainable recommendation has attracted increasing attentions, which can make the recommender system more transparent and improve user satisfactions by recommending products with useful explanations. However, existing methods trend to trade-off between the recommendation accuracy and the interpretability of recommendation results. In this manuscript, we propose Knowledge Enhanced Graph Neural Networks (KEGNN) for explainable recommendation. Semantic knowledge from the external knowledge base is leveraged into representation learning of three sides, respectively user, items and user-item interactions, and the knowledge enhanced semantic embedding are exploited to initialize the user/item entities and user-item relations of one constructed user behavior graph. We design a graph neural networks based user behavior learning and reasoning model to perform both semantic and relational knowledge propagation and reasoning over the user behavior graph for comprehensive understanding of user behaviors. On the top of comprehensive representations of users/items and user-item interactions, hierarchical neural collaborative filtering layers are developed for precise rating prediction, and one generation-mode and copy-mode combined generator is devised for human-like semantic explanation generation by integrating the copy mechanism into gated recurrent neural networks. Quantitative and qualitative results demonstrate the superiority of KEGNN over the state-of-art methods, and the explainability and interpretability of our method.

Significant progress has been made by joint entity and relation extraction methods, which directly generate the relation triplets and mitigate the issue of overlapping relations. However, previous models generate the entity-relation triplets solely from input sentences. Such information is insufficient to support the modeling of interactive information between entities and relations. In this paper, we define the features that provide mutual supports for entity and relation detection but can only be accessed at training time as privileged features for relation extraction, and devise two teacher models to exploit privileged entity and relation features, respectively. Meanwhile, we propose a novel contrastive student-teacher learning framework for joint extraction of entities and relations (STER), where a student network is encouraged to amalgamate privileged knowledge from two expert teacher networks that additionally utilize the privileged features, based on contrastive learning. Experiment results on three benchmark datasets (i.e., ADE, SciERC and CoNLL04) demonstrate that STER has robust superiority over competitors and sets state-of-the-art. For reproducibility, we will release the data and source code once the paper is accepted.

With the rapid growth of large-scale knowledge bases (KBs), knowledge base question answering (KBQA) has attracted increasing attention recently. Relation detection plays an important role in the KBQA system, which finds a compatible answer by analyzing the semantics of questions and querying and reasoning with multiple KB triples. Significant progress has been made by deep neural networks. However, existing methods often concern on detecting single-hop relation without path reasoning, and a few of these methods exploit the multihop relation reasoning, which involves the answer reasoning from the noisy and abundant relational paths in the KB. Meanwhile, the relatedness between question and answer candidates has received little attention and remains unsolved. This article proposes a novel knowledge-based reasoning network (KRN) for relation detection, including both single-hop relation and multihop relation. To address the semantic gap problem in question-answer interaction, we first learn attentive question representations according to the influence of answer aspects. Then, we learn the single-hop relation sequence through different levels of abstraction and adopt the KB entity and structure information to denoise the multihop relation detection task. Finally, we adopt a Siamese network to measure the similarity between question representation and relation representation for both single-hop and multihop relation KBQA tasks. We conduct experiments on two well-known benchmarks, SimpleQuestions and WebQSP, and the results show the superiority of our approach over the state-of-the-art models for both single-hop and multihop relation detection. Our model also achieves a significant improvement over existing methods on KBQA end task. Further analysis demonstrates the robustness and the applicability of the proposed approach.

Existing aspect-based/category sentiment analysis methods have shown great success in detecting sentiment polarity toward a given aspect in a sentence with supervised learning, where the training and inference stages share the same pre-defined set of aspects. However, in practice, the aspect categories are changing rather than keeping fixed over time. Dealing with unseen aspect categories is under-explored in existing methods. In this article, we formulate a new few-shot aspect category sentiment analysis (FSACSA) task, which aims to effectively predict the sentiment polarity of previously unseen aspect categories. To this end, we propose a novel Aspect-Focused Meta-Learning (AFML) framework that constructs aspect-aware and aspect-contrastive representations from external knowledge to match the target aspect with aspects in the training set. Concretely, we first construct two auxiliary contrastive sentences for a given sentence with the incorporation of external knowledge, enabling the learning of sentence representations with a better generalization. Then, we devise an aspect-focused induction network to leverage the contextual sentiment toward a given aspect to refine the label vectors. Furthermore, we employ the episode-based meta-learning algorithm to train the whole network, so as to learn to generalize to novel aspects. Extensive experiments on multiple real-life datasets show that our proposed AFML framework achieves the state-of-the-art results for the FSACSA task.

Pre-trained models have proved to be powerful in enhancing task-oriented dialog systems. However, current pre-training methods mainly focus on enhancing dialog understanding and generation tasks while neglecting the exploitation of dialog policy. In this paper, we propose GALAXY, a novel pre-trained dialog model that explicitly learns dialog policy from limited labeled dialogs and large-scale unlabeled dialog corpora via semi-supervised learning. Specifically, we introduce a dialog act prediction task for policy optimization during pre-training and employ a consistency regularization term to refine the learned representation with the help of unlabeled dialogs. We also implement a gating mechanism to weigh suitable unlabeled dialog samples. Empirical results show that GALAXY substantially improves the performance of task-oriented dialog systems, and achieves new state-of-the-art results on benchmark datasets: In-Car, MultiWOZ2.0 and MultiWOZ2.1, improving their end-to-end combined scores by 2.5, 5.3 and 5.5 points, respectively. We also show that GALAXY has a stronger few-shot ability than existing models under various low-resource settings. For reproducibility, we release the code and data at https://github.com/siat-nlp/GALAXY.

Bin Liang, Chenwei Lou, Xiang Li, Min Yang, Lin Gui, Yulan He, Wenjie Pei, Ruifeng Xu. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

Zero-shot stance detection (ZSSD) aims to detect the stance for an unseen target during the inference stage.In this paper, we propose a joint contrastive learning (JointCL) framework, which consists of stance contrastive learning and target-aware prototypical graph contrastive learning.Specifically, a stance contrastive learning strategy is employed to better generalize stance features for unseen targets.Further, we build a prototypical graph for each instance to learn the target-based representation, in which the prototypes are deployed as a bridge to share the graph structures between the known targets and the unseen ones.Then a novel target-aware prototypical graph contrastive learning strategy is devised to generalize the reasoning ability of target-based stance representations to the unseen targets.Extensive experiments on three benchmark datasets show that the proposed approach achieves state-ofthe-art performance in the ZSSD task 1 .

Pre-training methods with contrastive learning objectives have shown remarkable success in dialog understanding tasks. However, current contrastive learning solely considers the self-augmented dialog samples as positive samples and treats all other dialog samples as negative ones, which enforces dissimilar representations even for dialogs that are semantically related. In this paper, we propose SPACE-2, a tree-structured pre-trained conversation model, which learns dialog representations from limited labeled dialogs and large-scale unlabeled dialog corpora via semi-supervised contrastive pre-training. Concretely, we first define a general semantic tree structure (STS) to unify the inconsistent annotation schema across different dialog datasets, so that the rich structural information stored in all labeled data can be exploited. Then we propose a novel multi-view score function to increase the relevance of all possible dialogs that share similar STSs and only push away other completely different dialogs during supervised contrastive pre-training. To fully exploit unlabeled dialogs, a basic self-supervised contrastive loss is also added to refine the learned representations. Experiments show that our method can achieve new state-of-the-art results on the DialoGLUE benchmark consisting of seven datasets and four popular dialog understanding tasks.

The existing research efforts in Multimodal Sentiment Analysis (MSA) have focused on developing the expressive ability of neural networks to fuse information from different modalities. However, these approaches lack a mechanism to understand the complex relations within and across different modalities, since some sentiments may be scattered in different modalities. To this end, in this paper, we propose a novel hierarchical graph contrastive learning (HGraph-CL) framework for MSA, aiming to explore the intricate relations of intra- and inter-modal representations for sentiment extraction. Specifically, regarding the intra-modal level, we build a unimodal graph for each modality representation to account for the modality-specific sentiment implications. Based on it, a graph contrastive learning strategy is adopted to explore the potential relations based on unimodal graph augmentations. Furthermore, we construct a multimodal graph of each instance based on the unimodal graphs to grasp the sentiment relations between different modalities. Then, in light of the multimodal augmentation graphs, a graph contrastive learning strategy over the inter-modal level is proposed to ulteriorly seek the possible graph structures for precisely learning sentiment relations. This essentially allows the framework to understand the appropriate graph structures for learning intricate relations among different modalities. Experimental results on two benchmark datasets show that the proposed framework outperforms the state-of-the-art baselines in MSA.

This paper aims to improve the performance of text-to-SQL parsing by exploring the intrinsic uncertainties in the neural network based approaches (called SUN). From the data uncertainty perspective, it is indisputable that a single SQL can be learned from multiple semantically-equivalent questions. Different from previous methods that are limited to one-to-one mapping, we propose a data uncertainty constraint to explore the underlying complementary semantic information among multiple semantically-equivalent questions (many-to-one) and learn the robust feature representations with reduced spurious associations. In this way, we can reduce the sensitivity of the learned representations and improve the robustness of the parser. From the model uncertainty perspective, there is often structural information (dependence) among the weights of neural networks. To improve the generalizability and stability of neural text-to-SQL parsers, we propose a model uncertainty constraint to refine the query representations by enforcing the output representations of different perturbed encoding networks to be consistent with each other. Extensive experiments on five benchmark datasets demonstrate that our method significantly outperforms strong competitors and achieves new state-of-the-art results.

Few-shot relation classification aims to classify the relation type between two given entities in a sentence by training with a few labeled instances for each relation. However, most of existing models fail to distinguish multiple relations that co-exist in one sentence. This paper presents a novel dependency-aware prototype learning (DAPL) method for few-shot relation classification. Concretely, we utilize dependency trees and shortest dependency paths (SDP) as structural information to complement the contextualized representations of input sentences by using the dependency-aware embedding as attention inputs to learn attentive sentence representations. In addition, we introduce a gate controlled update mechanism to update the dependency-aware representations according to the output of each network layer. Extensive experiments on the FewRel dataset show that DAPL achieves substantially better performance than strong baselines. For reproducibility, we will release our code and data upon the publication of this paper at https://github.com/publicstaticvo/DAPL.

Deep neural models have become the mainstream in answer selection, yielding state-of-the-art performance. However, these models tend to rely on spurious correlations between prediction labels and input features, which in general suffer from robustness and generalization. In this paper, we propose a novel Spurious Correlation reduction method to improve the robustness of the neural ANswer selection models (SCAN) from the sample and feature perspectives by removing the feature dependencies and language biases in answer selection. First, from the sample perspective, we propose a feature decorrelation module by learning a weight for each instance at the training phase to remove the feature dependencies and reduce the spurious correlations without prior knowledge of such correlations. Second, from the feature perspective, we propose a feature debiasing module with contrastive learning to alleviate the negative language biases (spurious correlations) and further improve the robustness of the AS models. Experimental results on three benchmark datasets show that SCAN achieves substantial improvements over strong baselines. For reproducibility, we will release our code and data upon the publication of this paper.

Argument mining (AM) is a challenging task as it requires recognizing the complex argumentation structures involving multiple subtasks.To handle all subtasks of AM in an end-to-end fashion, previous works generally transform AM into a dependency parsing task.However, such methods largely require complex pre-and post-processing to realize the task transformation.In this paper, we investigate the endto-end AM task from a novel perspective by proposing a generative framework, in which the expected outputs of AM are framed as a simple target sequence.Then, we employ a pretrained sequence-to-sequence language model with a constrained pointer mechanism (CPM) to model the clues for all the subtasks of AM in the light of the target sequence.Furthermore, we devise a reconstructed positional encoding (RPE) to alleviate the order biases induced by the autoregressive generation paradigm.Experimental results show that our proposed framework achieves new state-of-the-art performance on two AM benchmarks. 1

This paper investigates the sentiment analysis task from a novel perspective by incorporating sentiment knowledge and eye movement into a graph architecture, aiming to draw the eye movement-based sentiment relationships for learning the sentiment expression of the context.To be specific, we first explore a linguistic probing eye movement paradigm to extract eye movement features based on the close relationship between linguistic features and the early and late processes of human reading behavior.Furthermore, to derive eye movement features with sentiment concepts, we devise a novel weighting strategy to integrate sentiment scores extracted from affective commonsense knowledge into eye movement features, called sentiment-eye movement weights.Then, the sentiment-eye movement weights are exploited to build the sentiment-eye movement guided graph (SEMGraph) model, so as to model the intricate sentiment relationships in the context.Experimental results on two sentiment analysis datasets with eye movement signals and three sentiment analysis datasets without eye movement signals show that the proposed SEM-Graph achieves state-of-the-art performance, and can also be directly generalized to those sentiment analysis datasets without eye movement signals.

Aspect Sentiment Triplet Extraction (ASTE) aims to extract the aspect terms along with the corresponding opinion terms and the expressed sentiments in the review, which is an important task in sentiment analysis.Previous research efforts generally address the ASTE task in an endto-end fashion through the table-filling formalization, in which the triplets are represented by a two-dimensional (2D) table of word-pair relations.Under this formalization, a term-level relation is decomposed into multiple independent word-level relations, which leads to relation inconsistency and boundary insensitivity in the face of multi-word aspect terms and opinion terms.To overcome these issues, we propose Boundary-Driven Table-Filling (BDTF), which represents each triplet as a relation region in the 2D table and transforms the ASTE task into detection and classification of relation regions.We also notice that the quality of the table representation greatly affects the performance of BDTF.Therefore, we develop an effective relation representation learning approach to learn the table representation, which can fully exploit both word-to-word interactions and relation-torelation interactions.Experiments on several public benchmarks show that the proposed approach achieves state-of-the-art performances 1 .

The importance of building text-to-SQL parsers which can be applied to new databases has long been acknowledged, and a critical step to achieve this goal is schema linking, i.e., properly recognizing mentions of unseen columns or tables when generating SQLs. In this work, we propose a novel framework to elicit relational structures from large-scale pre-trained language models (PLMs) via a probing procedure based on Poincaré distance metric, and use the induced relations to augment current graph-based parsers for better schema linking. Compared with commonly-used rule-based methods for schema linking, we found that probing relations can robustly capture semantic correspondences, even when surface forms of mentions and entities differ. Moreover, our probing procedure is entirely unsupervised and requires no additional parameters. Extensive experiments show that our framework sets new state-of-the-art performance on three benchmarks. We empirically verify that our probing procedure can indeed find desired relational structures through qualitative analysis.

Existing benchmark datasets for recommender systems (RS) either are created at a small scale or involve very limited forms of user feedback. RS models evaluated on such datasets often lack practical values for large-scale real-world applications. In this paper, we describe Tenrec, a novel and publicly available data collection for RS that records various user feedback from four different recommendation scenarios. To be specific, Tenrec has the following five characteristics: (1) it is large-scale, containing around 5 million users and 140 million interactions; (2) it has not only positive user feedback, but also true negative feedback (vs. one-class recommendation); (3) it contains overlapped users and items across four different scenarios; (4) it contains various types of user positive feedback, in forms of clicking, liking, sharing, and following, etc; (5) it contains additional features beyond the user IDs and item IDs. We verify Tenrec on ten diverse recommendation tasks by running several classical baseline models per task. Tenrec has the potential to become a useful benchmark dataset for a majority of popular recommendation tasks. Our source codes and datasets will be included in supplementary materials.

Sequential recommender systems (SRSs) have become a research hotspot recently due to its powerful ability in capturing users' dynamic preferences. The key idea behind SRSs is to model the sequential dependencies over the user-item interactions. However, we argue that users' preferences are not only determined by their view or purchase items but also affected by the item-providers with which users have interacted. For instance, in a short-video scenario, a user may click on a video because he/she is attracted to either the video content or simply the video-providers as the vloggers are his/her idols. Motivated by the above observations, in this paper, we propose IPSRec, a novel Item-Provider co-learning framework for Sequential Recommendation. Specifically, we propose two representation learning methods (single-steam and cross-stream) to learn comprehensive item and user representations based on the user's historical item sequence and provider sequence. Then, contrastive learning is employed to further enhance the user embeddings in a self-supervised manner, which treats the representations of a specific user learned from the item side as well as the item-provider side as the positive pair and treats the representations of different users in the batch as the negative samples. Extensive experiments on three real-world SRS datasets demonstrate that IPSRec achieves substantially better results than the strong competitors. For reproducibility, our code and data are available at https://github.com/siat-nlp/IPSRec.

Recently, pre-training methods have shown remarkable success in task-oriented dialog (TOD) systems. However, most existing pre-trained models for TOD focus on either dialog understanding or dialog generation, but not both. In this paper, we propose SPACE, a novel unified pre-trained dialog model learning from large-scale dialog corpora with limited annotations, which can be effectively fine-tuned on a wide range of downstream dialog tasks. Specifically, SPACE consists of four successive components in a single transformer to maintain a task-flow in TOD systems: (i) a dialog encoding module to encode dialog history, (ii) a dialog understanding module to extract semantic vectors from either user queries or system responses, (iii) a dialog policy module to generate a policy vector that contains high-level semantics of the response, and (iv) a dialog generation module to produce appropriate responses. We design a dedicated pre-training objective for each component. Concretely, we pre-train the dialog encoding module with span mask language modeling to learn contextualized dialog information. To capture the structured dialog semantics, we pre-train the dialog understanding module via a novel tree-induced semi-supervised contrastive learning objective with the help of extra dialog annotations. In addition, we pre-train the dialog policy module by minimizing the ℒ2 distance between its output policy vector and the semantic vector of the response for policy optimization. Finally, the dialog generation model is pre-trained by language modeling. Results show that SPACE achieves state-of-the-art performance on eight downstream dialog benchmarks, including intent prediction, dialog state tracking, and end-to-end dialog modeling. We also show that SPACE has a stronger few-shot ability than existing models under the low-resource setting.

In this paper, we study the semi-supervised text classification (SSTC) by exploring both labeled and extra unlabeled data. One of the most popular SSTC techniques is pseudo-labeling which assigns pseudo labels for unlabeled data via a teacher classifier trained on labeled data. These pseudo labeled data is then applied to train a student classifier. However, when the pseudo labels are inaccurate, the student classifier will learn from inaccurate data and get even worse performance than the teacher. To mitigate this issue, we propose a simple yet efficient pseudo-labeling framework called Dual Pseudo Supervision (DPS), which exploits the feedback signal from the student to guide the teacher to generate better pseudo labels. In particular, we alternately update the student based on the pseudo labeled data annotated by the teacher and optimize the teacher based on the student's performance via meta learning. In addition, we also design a consistency regularization term to further improve the stability of the teacher. With the above two strategies, the learned reliable teacher can provide more accurate pseudo-labels to the student and thus improve the overall performance of text classification. We conduct extensive experiments on three benchmark datasets (i.e., AG News, Yelp and Yahoo) to verify the effectiveness of our DPS method. Experimental results show that our approach achieves substantially better performance than the strong competitors. For reproducibility, we will release our code and data of this paper publicly at https://github.com/GRIT621/DPS.

Deep neural networks (DNN) based recommender models often require numerous parameters to achieve remarkable performance. However, this inevitably brings redundant neurons, a phenomenon referred to as over-parameterization. In this paper, we plan to exploit such redundancy phenomena for recommender systems (RS), and propose a top-N item recommendation framework called PCRec that leverages collaborative training of two recommender models of the same network structure, termed peer collaboration. We first introduce two criteria to identify the importance of parameters of a given recommender model. Then, we rejuvenate the unimportant parameters by copying parameters from its peer network. After such an operation and retraining, the original recommender model is endowed with more representation capacity by possessing more functional model parameters. To show its generality, we instantiate PCRec by using three well-known recommender models. We conduct extensive experiments on two real-world datasets, and show that PCRec yields significantly better performance than its counterpart with the same model (parameter) size.

Math Word Problems (MWP) aims to automatically solve mathematical questions given in texts. Previous studies tend to design complex models to capture additional information in the original text so as to enable the model to gain more comprehensive features. In this paper, we turn our attention in the opposite direction, and work on how to discard redundant features containing spurious correlations for MWP. To this end, we design an Expression Syntax Information Bottleneck method for MWP (called ESIB) based on variational information bottleneck, which extracts essential features of the expression syntax tree while filtering latent-specific redundancy containing syntax-irrelevant features. The key idea of ESIB is to encourage multiple models to predict the same expression syntax tree for different problem representations of the same problem by mutual learning so as to capture consistent information of expression syntax tree and discard latent-specific redundancy. To improve the generalization ability of the model and generate more diverse expressions, we design a self-distillation loss to encourage the model to rely more on the expression syntax information in the latent space. Experimental results on two large-scale benchmarks show that our model not only achieves state-of-the-art results but also generates more diverse solutions.

In this paper, we bridge the heterogeneity gap between different modalities and improve image-text retrieval by taking advantage of auxiliary image-to-text and text-to-image generative features with contrastive learning. Concretely, contrastive learning is devised to narrow the distance between the aligned image-text pairs and push apart the distance between the unaligned pairs from both inter- and intra-modality perspectives with the help of cross-modal retrieval features and auxiliary generative features. In addition, we devise a support-set regularization term to further improve contrastive learning by constraining the distance between each image/text and its corresponding cross-modal support-set information contained in the same semantic category. To evaluate the effectiveness of the proposed method, we conduct experiments on three benchmark datasets (i.e., MIRFLICKR-25K, NUS-WIDE, MS COCO). Experimental results show that our model significantly outperforms the strong baselines for cross-modal image-text retrieval. For reproducibility, we submit the code and data publicly at: \urlhttps://github.com/Hambaobao/CRCGS.

Zero-shot stance detection (ZSSD) is challenging as it requires detecting the stance of previously unseen targets during the inference stage. Being able to detect the target-related transferable stance features from the training data is arguably an important step in ZSSD. Generally speaking, stance features can be grouped into target-invariant and target-specific categories. Target-invariant stance features carry the same stance regardless of the targets they are associated with. On the contrary, target-specific stance features only co-occur with certain targets. As such, it is important to distinguish these two types of stance features when learning stance features of unseen targets. To this end, in this paper, we revisit ZSSD from a novel perspective by developing an effective approach to distinguish the types (target-invariant/-specific) of stance features, so as to better learn transferable stance features. To be specific, inspired by self-supervised learning, we frame the stance-feature-type identification as a pretext task in ZSSD. Furthermore, we devise a novel hierarchical contrastive learning strategy to capture the correlation and difference between target-invariant and -specific features and further among different stance labels. This essentially allows the model to exploit transferable stance features more effectively for representing the stance of previously unseen targets. Extensive experiments on three benchmark datasets show that the proposed framework achieves the state-of-the-art performance in ZSSD.

Continual relation extraction (CRE) is an important task of continual learning, which aims to learn incessantly emerging new relations between entities from texts. To avoid catastrophically forgetting old relations, some existing research efforts have focused on exploring memory replayed methods by storing typical historical learned instances or embedding all observed relations as prototypes in the episodic memory and replaying them in the subsequent training process. However, they generally fail to exploit the relation knowledge contained in the pre-trained language model (PLM), which could provide enlightening information to the representations of new relations from the known ones. To this end, we investigate the CRE from a novel perspective by generating knowledge-infused relation prototypes to leverage the relational knowledge from PLM with prompt tuning. Specifically, based on the typical samples collected from the historical learned instances with K-means algorithm, we devise novel relational knowledge-infused prompts to elicit relational knowledge from PLM for generating knowledge-infused relation prototypes. Then the prototypes are used to refine the typical examples embedding and calculate the stability-plasticity balance score for adjusting the memory replayed progress. The experimental results show that our method outperforms the state-of-the-art baseline models in CRE. The further extensive analysis presents that the proposed method is robust to memory size, task order, length of the task sequence, and the number of training instances.

Recently, semisupervised feature selection has gained more attention in many real applications due to the high cost of obtaining labeled data. However, existing methods cannot solve the "multimodality" problem that samples in some classes lie in several separate clusters. To solve the multimodality problem, this article proposes a new feature selection method for semisupervised task, namely, semisupervised structured manifold learning (SSML). The new method learns a new structured graph which consists of more clusters than the known classes. Meanwhile, we propose to exploit the submanifold in both labeled data and unlabeled data by consuming the nearest neighbors of each object in both labeled and unlabeled objects. An iterative optimization algorithm is proposed to solve the new model. A series of experiments was conducted on both synthetic and real-world datasets and the experimental results verify the ability of the new method to solve the multimodality problem and its superior performance compared with the state-of-the-art methods.

In big data time, selecting informative features has become an urgent need. However, due to the huge cost of obtaining enough labeled data for supervised tasks, researchers have turned their attention to semisupervised learning, which exploits both labeled and unlabeled data. In this article, we propose a sparse discriminative semisupervised feature selection (SDSSFS) method. In this method, the$\epsilon $-dragging technique for the supervised task is extended to the semisupervised task, which is used to enlarge the distance between classes in order to obtain a discriminative solution. The flexible$\ell _{2,p}$norm is implicitly used as regularization in the new model. Therefore, we can obtain a more sparse solution by setting smaller$p$. An iterative method is proposed to simultaneously learn the regression coefficients and$\epsilon $-dragging matrix and predicting the unknown class labels. Experimental results on ten real-world datasets show the superiority of our proposed method.

One of the main challenges in multi-source domain adaptation is how to reduce the domain discrepancy between each source domain and a target domain, and then evaluate the domain relevance to determine how much knowledge should be transferred from different source domains to the target domain. However, most prior approaches barely consider both discrepancies and relevance among domains. In this paper, we propose an algorithm, called Iterative Refinement based on Feature Selection and the Wasserstein distance (IRFSW), to solve semi-supervised domain adaptation with multiple sources. Specifically, IRFSW aims to explore both the discrepancies and relevance among domains in an iterative learning procedure, which gradually refines the learning performance until the algorithm stops. In each iteration, for each source domain and the target domain, we develop a sparse model to select features in which the domain discrepancy and training loss are reduced simultaneously. Then a classifier is constructed with the selected features of the source and labeled target data. After that, we exploit optimal transport over the selected features to calculate the transferred weights. The weight values are taken as the ensemble weights to combine the learned classifiers to control the amount of knowledge transferred from source domains to the target domain. Experimental results validate the effectiveness of the proposed method.

Answer selection, which is involved in many natural language processing applications, such as dialog systems and question answering (QA), is an important yet challenging task in practice, since conventional methods typically suffer from the issues of ignoring diverse real-world background knowledge. In this article, we extensively investigate approaches to enhancing the answer selection model with external knowledge from knowledge graph (KG). First, we present a context-knowledge interaction learning framework, Knowledge-aware Neural Network, which learns the QA sentence representations by considering a tight interaction with the external knowledge from KG and the textual information. Then, we develop two kinds of knowledge-aware attention mechanism to summarize both the context-based and knowledge-based interactions between questions and answers. To handle the diversity and complexity of KG information, we further propose a Contextualized Knowledge-aware Attentive Neural Network, which improves the knowledge representation learning with structure information via a customized Graph Convolutional Network and comprehensively learns context-based and knowledge-based sentence representation via the multi-view knowledge-aware attention mechanism. We evaluate our method on four widely used benchmark QA datasets, including WikiQA, TREC QA, InsuranceQA, and Yahoo QA. Results verify the benefits of incorporating external knowledge from KG and show the robust superiority and extensive applicability of our method.

Sequential recommender systems (SRS) have become a research hotspot in recent studies. Because of the requirement in capturing user's dynamic interests, sequential neural network based recommender models often need to be stacked with more hidden layers (e.g., up to 100 layers) compared with standard collaborative filtering methods. However, the high network latency has become the main obstacle when deploying very deep recommender models into a production environment. In this paper, we argue that the typical prediction framework that treats all users equally during the inference phase is inefficient in running time, as well as sub-optimal in accuracy. To resolve such an issue, we present SkipRec, an adaptive inference framework by learning to skip inactive hidden layers on a per-user basis. Specifically, we devise a policy network to automatically determine which layers should be retained and which layers are allowed to be skipped, so as to achieve user-specific decisions. To derive the optimal skipping policy, we propose using gumbel softmax and reinforcement learning to solve the non-differentiable problem during backpropagation. We perform extensive experiments on three real-world recommendation datasets, and demonstrate that SkipRec attains comparable or better accuracy with much less inference time.

Eye-catching headlines function as the first device to trigger more clicks, bringing reciprocal effect between producers and viewers. Producers can obtain more traffic and profits, and readers can have access to outstanding articles. When generating attractive headlines, it is important to not only capture the attractive content but also follow an eye-catching writtenstyle. In this paper, we propose a Disentanglement-based Attractive Headline Generator (DAHG) that generates headline which captures the attractive content following the attractive style. Concretely, we first devise a disentanglement module to divide the style and content of an attractive prototype headline into latent spaces, with two auxiliary constraints to ensure the two spaces are indeed disentangled. The latent content information is then used to further polish the document representation and help capture the salient part. Finally, the generator takes the polished document as input to generate headline under the guidance of the attractive style. Extensive experiments on the public Kuaibao dataset show that DAHG achieves state-of-the-art performance. Human evaluation also demonstrates that DAHG triggers 22% more clicks than existing models.

In this paper, we propose a Meta Cooperative Learning (MCL) framework for task-oriented dialog systems (TDSs). Our model consists of an auxiliary KB reasoning task for learning meta KB knowledge, an auxiliary dialogue reasoning task for learning dialogue patterns, and a TDS task (primary task) that aims at not only retrieving accurate entities from KB but also generating natural responses, which are coordinated to achieve collective success in both retrieving accurate KB entities and generating human-like responses via meta learning. Concretely, the dialog generation model amalgamates complementary meta KB and dialog knowledge from two novel auxiliary reasoning tasks that together provide integrated guidance to build a high-quality TDS by adding regularization terms to force primary network to produce similar results to auxiliary networks. While MCL automatically learns appropriate labels for the two auxiliary reasoning tasks from the primary task, without requiring access to any further data. The key idea behind MCL is to use the performance of the primary task, which is trained alongside the auxiliary tasks in one iteration, to improve the auxiliary labels for the next iteration with meta learning. Experimental results on three benchmark datasets show that MCL can generate higher quality responses compared to several strong baselines in terms of both automatic and human evaluations. Code to reproduce the results in this paper is available at: https://github.com/siat-nlp/MCL.

Visual storytelling is a task of creating a short story based on photo streams. Different from visual captions, stories contain not only factual descriptions, but also imaginary concepts that do not appear in the images. In this paper, we propose a novel imagine-reason-write generation framework (IRW) for visual storytelling, inspired by the logic of humans when they write the story. First, an imagine module is leveraged to learn the imaginative storyline explicitly, improving the coherence and reasonability of the generated story. Second, we employ a reason module to fully exploit the external knowledge (commonsense knowledge base) and task-specific knowledge (scene graph and event graph) with relational reasoning method based on the storyline. In this way, we can effectively capture the most informative commonsense and visual relationships among objects in images, which enhances the diversity and informativeness of the generated story. Finally, we integrate the imaginary concepts and relational knowledge to generate human-like story based on the original semantics of images. Extensive experiments on a benchmark dataset (i.e., VIST) demonstrate that the proposed IRW framework significantly outperforms the state-of-the-art methods across multiple evaluation metrics.

Junhao Liu, Zhen Hai, Min Yang, Lidong Bing. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

With the ever-increasing size of legal cases in China, relevant legal case retrieval given a user query has attracted considerable attention. Conventional keyword-based retrieval systems look for matching cases that contain one or more words specified by the user. However, keyword search is sharply focused on finding the exact terms specified in the query, making the retrieval systems miss many relevant documents. In addition, it is difficult for new users to identify appropriate keywords for accurate legal case retrieval. In this paper, we develop a novel Chinese legal case retrieval system (called CLC-RS), which improves the quality of semantic search with natural language queries in the legal domain. CLC-RS performs legal case retrieval in a two-stage fashion. First, we employ a classic token-based ranking method to efficiently reduce the solution space, returning a subset of candidate legal cases. Then, we deploy a novel masked language ranking model to re-rank the candidate legal cases. The experimental results show that the proposed system is both efficient and effective, providing a practical information retrieval (IR) system for retrieving Chinese legal cases. The web site for the developed CLC-RS system is available at: https://www.delilegal.com/.

Most existing aspect-based sentiment analysis (ABSA) research efforts are devoted to extracting the aspect-dependent sentiment features from the sentence towards the given aspect. However, it is observed that about 60% of the testing aspects in commonly used public datasets are unknown to the training set. That is, some sentiment features carry the same polarity regardless of the aspects they are associated with (aspect-invariant sentiment), which props up the high accuracy of existing ABSA models when inevitably inferring sentiment polarities for those unknown testing aspects. Therefore, in this paper, we revisit ABSA from a novel perspective by deploying a novel supervised contrastive learning framework to leverage the correlation and difference among different sentiment polarities and between different sentiment patterns (aspect-invariant/-dependent). This allows improving the sentiment prediction for (unknown) testing aspects in the light of distinguishing the roles of valuable sentiment features. Experimental results on 5 benchmark datasets show that our proposed approach substantially outperforms state-of-the-art baselines in ABSA. We further extend existing neural network-based ABSA models with our proposed framework and achieve improved performance.

Auxiliary information from multiple sources has been demonstrated to be effective in zeroshot fine-grained entity typing (ZFET).However, there lacks a comprehensive understanding about how to make better use of the existing information sources and how they affect the performance of ZFET.In this paper, we empirically study three kinds of auxiliary information: context consistency, type hierarchy and background knowledge (e.g., prototypes and descriptions) of types, and propose a multi-source fusion model (MSF) targeting these sources.The performance obtains up to 11.42% and 22.84% absolute gains over stateof-the-art baselines on BBN and Wiki respectively with regard to macro F1 scores.More importantly, we further discuss the characteristics, merits and demerits of each information source and provide an intuitive understanding of the complementarity among them.

Argument pair extraction (APE) aims to extract interactive argument pairs from two passages of a discussion.Previous work studied this task in the context of peer review and rebuttal, and decomposed it into a sequence labeling task and a sentence relation classification task.However, despite the promising performance, such an approach obtains the argument pairs implicitly by the two decomposed tasks, lacking explicitly modeling of the argument-level interactions between argument pairs.In this paper, we tackle the APE task by a mutual guidance framework, which could utilize the information of an argument in one passage to guide the identification of arguments that can form pairs with it in another passage.In this manner, two passages can mutually guide each other in the process of APE.Furthermore, we propose an inter-sentence relation graph to effectively model the interrelations between two sentences and thus facilitates the extraction of argument pairs.Our proposed method can better represent the holistic argument-level semantics and thus explicitly capture the complex correlations between argument pairs.Experimental results show that our approach significantly outperforms the current state-of-the-art model.

In this paper, we investigate the Aspect Category Sentiment Analysis (ACSA) task from a novel perspective by exploring a Beta Distribution guided aspect-aware graph construction based on external knowledge.That is, we are no longer entangled about how to laboriously search the sentiment clues for coarsegrained aspects from the context, but how to preferably find the words highly sentimentrelated to the aspects in the context and determine their importance based on the public knowledge base, so as to naturally learn the aspect-related contextual sentiment dependencies with these words in ACSA.To be specific, we first regard each aspect as a pivot to derive aspect-aware words that are highly related to the aspect from external affective commonsense knowledge.Then, we employ Beta Distribution to educe the aspect-aware weight, which reflects the importance to the aspect, for each aspect-aware word.Afterward, the aspect-aware words are served as the substitutes of the coarse-grained aspect to construct graphs for leveraging the aspectrelated contextual sentiment dependencies in ACSA.Experiments on 6 benchmark datasets show that our approach significantly outperforms the state-of-the-art baseline methods.

Aspect term extraction aims to extract aspect terms from a review sentence that users have expressed opinions on.One of the remaining challenges for aspect term extraction resides in the lack of sufficient annotated data.While self-training is potentially an effective method to address this issue, the pseudo-labels it yields on unlabeled data could induce noise.In this paper, we use two means to alleviate the noise in the pseudo-labels.One is that inspired by the curriculum learning, we refine the conventional self-training to progressive self-training.Specifically, the base model infers pseudo-labels on a progressive subset at each iteration, where samples in the subset become harder and more numerous as the iteration proceeds.The other is that we use a discriminator to filter the noisy pseudo-labels.Experimental results on four SemEval datasets show that our model significantly outperforms the previous baselines and achieves state-ofthe-art performance.

Sarcasm is a peculiar form and sophisticated linguistic act to express the incongruity of someone's implied sentiment expression, which is a pervasive phenomenon in social media platforms. Compared with sarcasm detection purely on texts, multi-modal sarcasm detection is more adapted to the rapidly growing social media platforms, where people are interested in creating multi-modal messages. When focusing on the multi-modal sarcasm detection for tweets consisting of texts and images on Twitter, the significant clue of improving the performance of multi-modal sarcasm detection evolves into how to determine the incongruity relations between texts and images. In this paper, we investigate multi-modal sarcasm detection from a novel perspective, so as to determine the sentiment inconsistencies within a certain modality and across different modalities by constructing heterogeneous in-modal and cross-modal graphs (InCrossMGs) for each multi-modal example. Based on it, we explore an interactive graph convolution network (GCN) structure to jointly and interactively learn the incongruity relations of in-modal and cross-modal graphs for determining the significant clues in sarcasm detection. Experimental results demonstrate that our proposed model achieves state-of-the-art performance in multi-modal sarcasm detection.

Lifelong learning capabilities are crucial for sentiment classifiers to process continuous streams of opinioned information on the Web. However, performing lifelong learning is non-trivial for deep neural networks as continually training of incrementally available information inevitably results in catastrophic forgetting or interference. In this paper, we propose a novel i terative network p runing with uncertainty r egularization method for l ifelong s entiment classification (IPRLS), which leverages the principles of network pruning and weight regularization. By performing network pruning with uncertainty regularization in an iterative manner, IPRLS can adapt a single BERT model to work with continuously arriving data from multiple domains while avoiding catastrophic forgetting and interference. Specifically, we leverage an iterative pruning method to remove redundant parameters in large deep networks so that the freed-up space can then be employed to learn new tasks, tackling the catastrophic forgetting problem. Instead of keeping the old-tasks fixed when learning new tasks, we also use an uncertainty regularization based on the Bayesian online learning framework to constrain the update of old tasks weights in BERT, which enables positive backward transfer, i.e. learning new tasks improves performance on past tasks while protecting old knowledge from being lost. In addition, we propose a task-specific low-dimensional residual function in parallel to each layer of BERT, which makes IPRLS less prone to losing the knowledge saved in the base BERT network when learning a new task. Extensive experiments on 16 popular review corpora demonstrate that the proposed IPRLS method significantly outperforms the strong baselines for lifelong sentiment classification. For reproducibility, we submit the code and data at: \urlhttps://github.com/siat-nlp/IPRLS .

Recent AI research has witnessed increasing interests in automatically designing the architecture of deep neural networks, which is coined as neural architecture search (NAS). The automatically searched network architectures via NAS methods have outperformed manually designed architectures on some NLP tasks. However, training a large number of model configurations for efficient NAS is computationally expensive, creating a substantial barrier for applying NAS methods in real-life applications. In this paper, we propose to accelerate neural architecture search for natural language processing based on knowledge distillation (called KD-NAS). Specifically, instead of searching the optimal network architecture on the validation set conditioned on the optimal network weights on the training set, we learn the optimal network by minimizing the knowledge loss transferred from a pre-trained teacher network to the searching network based on Earth Mover's Distance (EMD). Experiments on five datasets show that our method achieves promising performance compared to strong competitors on both accuracy and searching speed. For reproducibility, we submit the code at: https://github.com/lxk00/KD-NAS-EMD.

Deep learning has brought great progress for the sequential recommendation (SR) tasks. With advanced network architectures, sequential recommender models can be stacked with many hidden layers, e.g., up to 100 layers on real-world recommendation datasets. Training such a deep network is difficult because it can be computationally very expensive and takes much longer time, especially in situations where there are tens of billions of user-item interactions. To deal with such a challenge, we present StackRec, a simple, yet very effective and efficient training framework for deep SR models by iterative layer stacking. Specifically, we first offer an important insight that hidden layers/blocks in a well-trained deep SR model have very similar distributions. Enlightened by this, we propose the stacking operation on the pre-trained layers/blocks to transfer knowledge from a shallower model to a deep model, then we perform iterative stacking so as to yield a much deeper but easier-to-train SR model. We validate the performance of StackRec by instantiating it with four state-of-the-art SR models in three practical scenarios with real-world datasets. Extensive experiments show that StackRec achieves not only comparable performance, but also substantial acceleration in training time, compared to SR models that are trained from scratch. Codes are available at https://github.com/wangjiachun0426/StackRec.

In this paper, we propose a novel abstractive text summarization method with hierarchical multi-scale abstraction modeling and dynamic memory (called MADY). First, we propose a hierarchical multi-scale abstraction modeling method to capture the temporal dependencies of the document from multiple hierarchical levels of abstraction, which mimics the process of how human beings comprehend an article by learning fine timescales for low-level abstraction layers and coarse timescales for high-level abstraction layers. By applying this adaptive updating mechanism, the high-level abstraction layers are updated less frequently and expected to remember the long-term dependency better than the low-level abstraction layer. Second, we propose a dynamic key-value memory-augmented attention network to keep track of the attention history and comprehensive context information for the salient facets within the input document. In this way, our model can avoid generating repetitive words and faultiness summaries. Extensive experiments on two widely-used datasets demonstrate the effectiveness of the proposed MADY model in terms of both automatic evaluation and human evaluation. For reproducibility, we submit the code and data at: https://github.com/siat-nlp/MADY.git.

Target plays an essential role in stance detection of an opinionated review/claim, since the stance expressed in the text often depends on the target. In practice, we need to deal with targets unseen in the annotated training data. As such, detecting stance for an unknown or unseen target is an important research problem. This paper presents a novel approach that automatically identifies and adapts the target-dependent and target-independent roles that a word plays with respect to a specific target in stance expressions, so as to achieve cross-target stance detection. More concretely, we explore a novel solution of constructing heterogeneous target-adaptive pragmatics dependency graphs (TPDG) for each sentence towards a given target. An in-target graph is constructed to produce inherent pragmatics dependencies of words for a distinct target. In addition, another cross-target graph is constructed to develop the versatility of words across all targets for boosting the learning of dominant word-level stance expressions available to an unknown target. A novel graph-aware model with interactive Graphical Convolutional Network (GCN) blocks is developed to derive the target-adaptive graph representation of the context for stance detection. The experimental results on a number of benchmark datasets show that our proposed model outperforms state-of-the-art methods in cross-target stance detection.

This work takes the lead to study the aspect-level sentiment classificationin the domain adaptation scenario. Given a document of any domains, the model needs to figure out the sentiments with respect to fine-grained aspects in the documents. Two main challenges exist in this problem. One is to build a robust document modeling across domains; the other is to mine the domain-specific aspects and make use of the sentiment lexicon. In this paper, we propose a novel approach Neural Attentive model for cross-domain Aspect-level sentiment CLassification (NAACL), which leverages the benefits of the supervised deep neural network as well as the unsupervised probabilistic generative model to strengthen the representation learning. NAACL jointly learns two tasks: (i) a domain classifier, working on documents in both the source and target domains to recognize the domain information of input texts and transfer knowledge from the source domain to the target domain. In particular, a weakly supervised Latent Dirichlet Allocation model (wsLDA) is proposed to learn the domain-specificaspectandsentiment lexiconrepresentations that are then used to calculate the aspect/lexicon-aware document representations via a multi-view attention mechanism; (ii) an aspect-level sentiment classifier, sharing the document modeling with the domain classifier. It makes use of the domain classification results and the aspect/sentiment-aware document representations to classify the aspect-level sentiment of the document in domain adaptation scenario. NAACL is evaluated on both English and Chinese datasets with the out-of-domain as well as in-domain setups. Quantitatively, the experiments demonstrate that NAACL has robust superiority over the compared methods in terms of classification accuracy and F1 score. The qualitative evaluation also shows that the proposed model is capable of reasonably paying attention to those words that are important to judge the sentiment polarity of the input text given an aspect.

A protein complex is a group of associated polypeptide chains which plays essential roles in the biological process. Given a graph representing protein-protein interactions (PPI) network, it is critical but non-trivial to detect protein complexes, the subsets of proteins that are tightly coupled, from it. Network embedding is a technique to learn low-dimensional representations of vertices in networks. It has been proved quite useful for community detection in social networks in recent years. However, unlike social networks, PPI network does not contain rich metadata, so that existing network embedding methods cannot fully capture the network structure of PPI to improve the effect of protein complexes detection significantly. We propose a semi-supervised network embedding model by adopting graph convolutional networks to detect densely connected subgraphs effectively. We compare the performance of our model with state-of-the-art approaches on three popular PPI networks with various data sizes and densities. The experimental results show that our approach significantly outperforms other approaches on all three PPI networks.

Video question generation is a challenging task in visual information retrieval, which generates questions given a sequence of video frames. The existing methods mainly tackle the problem of single-turn video question generation, but single-turn conversation usually can't meet the needs of video information acquisition. In this paper, we propose a new framework for single-turn VQG, which introduces attention mechanism to process inference of dialog history. And we introduce selection mechanism to choose from the candidate questions generated by each round of dialog history. In the framework, we leverage a recent video question answering model to predict the answer to the generated question and adopt the answer quality as rewards to fine-tune our model based on a reinforced learning mechanism. We also introduce a new task of multi-turn video question generation (M-VQG), which is generating multiple questions based on dialog history and video information to build conversation step by step. Our method achieves the state-of-the-art performance of the single-turn VQG task on two large-scale datasets, YouTube-Clips and TACoS-MultiLevel, and provides a baseline approach for M-VQG task.

Cross-modal image-text retrieval has emerged as a challenging task that requires the multimedia system to bridge the heterogeneity gap between different modalities. In this paper, we take full advantage of image-to-text and text-to-image generation models to improve the performance of the cross-modal image-text retrieval model by incorporating the text-grounded and image-grounded generative features into the cross-modal common space with a “Two-Teacher One-Student” learning framework. In addition, a dual regularizer network is designed to distinguish the mismatched image-text pairs from the matched ones. In this way, we can capture the fine-grained correspondence between modalities and distinguish the best-retrieved result from a candidate set. Extensive experiments on three benchmark datasets (i.e., MIRFLICKR-25K, NUS-WIDE, and MS COCO) show that our model can achieve state-of-the-art cross-modal retrieval results. In particular, our model improves the image-to-text and text-to-image retrieval accuracy by more than 22% over the best competitors on the MS COCO dataset.

Knowledge graphs (KG) often encounter knowledge incompleteness. The path reasoning that predicts the unknown path relation between pairwise entities based on existing facts is one of the most promising approaches to the knowledge graph completion. However, most conventional path reasoning methods exclusively consider the entity description included in fact triples, ignoring both the type information of entities and the interaction between different semantic representations. In this study, we propose a novel method, Type-aware Attentive Path Reasoning (TAPR), to complete the knowledge graph by simultaneously considering KG structural information, textual information, and type information. More specifically, we first leverage types to enrich the representational learning of entities and relationships. Next, we describe a type-level attention to select the most relevant type of given entity in a specific triple without any predefined rules or patterns to reduce the impact of noisy types. After learning the distributed representation of all paths, path-level attention assigns different weights to paths, from which relations among entity pairs are calculated. We conduct a series of experiments on a real-world dataset to demonstrate the effectiveness of TAPR. Experimental results show that our method significantly outperforms all baselines on link prediction and entity prediction tasks.

Open-domain dialog generation, which is a crucial component of artificial intelligence, is an essential and challenging problem. In this article, we present a personalized dialog system, which leverages the advantages of multitask learning and reinforcement learning for personalized dialogue generation (MRPDG). Specifically, MRPDG consists of two subtasks: 1) an author profiling module that recognizes user characteristics from the input sentence (auxiliary task) and 2) a personalized dialog generation system that generates informative, grammatical, and coherent responses with reinforcement learning algorithms (primary task). Three kinds of rewards are proposed to generate high-quality conversations. We investigate the effectiveness of three widely used reinforcement learning methods [i.e., Q-learning, policy gradient, and actor-critic (AC) algorithm] in a personalized dialog generation system and demonstrate that the AC algorithm achieves the best results on the underlying framework. Comprehensive experiments are conducted to evaluate the performance of the proposed model on two real-life data sets. Experimental results illustrate that MRPDG is able to produce high-quality personalized dialogs for users with different characteristics. Quantitatively, the proposed model can achieve better performance than the compared methods across different evaluation metrics, such as the human evaluation, BiLingual Evaluation Understudy (BLEU), and perplexity.

Developing an abstractive text summarization (ATS) system that is capable of generating concise, appropriate, and plausible summaries for the source documents is a long-term goal of artificial intelligence (AI). Recent advances in ATS are overwhelmingly contributed by deep learning techniques, which have taken the state-of-the-art of ATS to a new level. Despite the significant success of previous methods, generating high-quality and human-like abstractive summaries remains a challenge in practice. The human reading cognition, which is essential for reading comprehension and logical thinking, is still relatively new territory and underexplored in deep neural networks. In this article, we propose a novel Hierarchical Human-like deep neural network for ATS (HH-ATS), inspired by the process of how humans comprehend an article and write the corresponding summary. Specifically, HH-ATS is composed of three primary components (i.e., a knowledge-aware hierarchical attention module, a multitask learning module, and a dual discriminator generative adversarial network), which mimic the three stages of human reading cognition (i.e., rough reading, active reading, and postediting). Experimental results on two benchmark data sets (CNN/Daily Mail and Gigaword) demonstrate that HH-ATS consistently and substantially outperforms the compared methods.

Real-time event summarization (RES) aims at extracting a handful of document updates from an overwhelming document stream as the real-time event summary that tracks and summarizes the evolving event of interest. It has been attracting much attention, especially with the growth of streaming applications. Despite the effectiveness of previous studies, obtaining relevant, nonredundant, and timely event summaries remains challenging in real-life applications. This study proposes an effective Hybrid learning model for RES (HRES), which attempts to resolve all three challenges (i.e., nonredundancy, relevance, and timeliness) of RES in a unified framework. The main idea is to: 1) exploit the factual background knowledge from the knowledge base (KB) to capture the informative knowledge and implicit information from the input document/query for better text matching; 2) design a memory network to memorize the input facts temporally from the historical document stream and avoid pushing redundant facts in subsequent timesteps; 3) leverage relevance prediction as an auxiliary task to strengthen the document modeling and help to extract relevant documents; and 4) consider both historical dependencies and future uncertainty of the real-time document stream by exploiting the reinforcement learning technique. Extensive experiments demonstrate that HRES has robust superiority over competitors and gains the state-of-the-art results.

Community question answering (CQA) gains increasing popularity in both academy and industry recently. However, the redundancy and lengthiness issues of crowdsourced answers limit the performance of answer selection and lead to reading difficulties and misunderstandings for community users. To solve these problems, we tackle the tasks of answer selection and answer summary generation in CQA with a novel joint learning model. Specifically, we design a question-driven pointer-generator network, which exploits the correlation information between question-answer pairs to aid in attending the essential information when generating answer summaries. Meanwhile, we leverage the answer summaries to alleviate noise in original lengthy answers when ranking the relevancy degrees of question-answer pairs. In addition, we construct a new large-scale CQA corpus, WikiHowQA, which contains long answers for answer selection as well as reference summaries for answer summarization. The experimental results show that the joint learning method can effectively address the answer redundancy issue in CQA and achieves state-of-the-art results on both answer selection and text summarization tasks. Furthermore, the proposed model is shown to be of great transferring ability and applicability for resource-poor CQA tasks, which lack of reference answer summaries.

Image captioning is usually built on either generation-based or retrieval-based approaches. Both ways have certain strengths but suffer from their own limitations. In this paper, we propose an Interactive Dual Generative Adversarial Network (IDGAN) for image captioning, which mutually combines the retrieval-based and generation-based methods to learn a better image captioning ensemble. IDGAN consists of two generators and two discriminators, where the generation- and retrieval-based generators mutually benefit from each other's complementary targets that are learned from two dual adversarial discriminators. Specifically, the generation- and retrieval-based generators provide improved synthetic and retrieved candidate captions with informative feedback signals from the two respective discriminators that are trained to distinguish the generated captions from the true captions and assign top rankings to true captions respectively, thus featuring the merits of both retrieval-based and generation-based approaches. Extensive experiments on MSCOCO dataset demonstrate that the proposed IDGAN model significantly outperforms the compared methods for image captioning.

Automatic short answer scoring (ASAS) is a research subject of intelligent education, which is a hot field of natural language understanding. Many experiments have confirmed that the ASAS system is not good enough, because its performance is limited by the training data. Focusing on the problem, we propose MDA-ASAS, multiple data augmentation strategies for improving performance on automatic short answer scoring. MDA-ASAS is designed to learn language representation enhanced by data augmentation strategies, which includes back-translation, correct answer as reference answer, and swap content. We argue that external knowledge has a profound impact on the ASAS process. Meanwhile, the Bidirectional Encoder Representations from Transformers (BERT) model has been shown to be effective for improving many natural language processing tasks, which acquires more semantic, grammatical and other features in large amounts of unsupervised data, and actually adds external knowledge. Combining with the latest BERT model, our experimental results on the ASAS dataset show that MDA-ASAS brings a significant gain over state-of-art. We also perform extensive ablation studies and suggest parameters for practical use.

Neural network models usually suffer from the challenge of incorporating commonsense knowledge into the open-domain dialogue systems. In this paper, we propose a novel knowledge-aware dialogue generation model (called TransDG), which transfers question representation and knowledge matching abilities from knowledge base question answering (KBQA) task to facilitate the utterance understanding and factual knowledge selection for dialogue generation. In addition, we propose a response guiding attention and a multi-step decoding strategy to steer our model to focus on relevant features for response generation. Experiments on two benchmark datasets demonstrate that our model has robust superiority over compared methods in generating informative and fluent dialogues. Our code is available at https://github.com/siat-nlp/TransDG.

Previous approaches for video summarization mainly concentrate on finding the most diverse and representative visual contents as video summary without considering the user's preference. This paper addresses the task of query-focused video summarization, which takes user's query and a long video as inputs and aims to generate a query-focused video summary. In this paper, we consider the task as a problem of computing similarity between video shots and query. To this end, we propose a method, named Convolutional Hierarchical Attention Network (CHAN), which consists of two parts: feature encoding network and query-relevance computing module. In the encoding network, we employ a convolutional network with local self-attention mechanism and query-aware global attention mechanism to learns visual information of each shot. The encoded features will be sent to query-relevance computing module to generate query-focused video summary. Extensive experiments on the benchmark dataset demonstrate the competitive performance and show the effectiveness of our approach.

We study the community question answering (CQA) problem that emerges with the advent of numerous community forums in the recent past. The task of finding appropriate answers to questions from informative but noisy crowdsourced answers is important yet challenging in practice. We present an Attentive User-engaged Adversarial Neural Network (AUANN), which interactively learns the context information of questions and answers, and enhances user engagement with the CQA task. A novel attentive mechanism is incorporated to model the semantic internal and external relations among questions, answers and user contexts. To handle the noise issue caused by introducing user context, we design a two-step denoise mechanism, including a coarse-grained selection process by similarity measurement, and a fine-grained selection process by applying an adversarial training module. We evaluate the proposed method on large-scale real-world datasets SemEval-2016 and SemEval-2017. Experimental results verify the benefits of incorporating user information, and show that our proposed model significantly outperforms the state-of-the-art methods.

Real-time event summarization is an essential task in natural language processing and information retrieval areas. Despite the progress of previous work, generating relevant, non-redundant, and timely event summaries remains challenging in practice. In this paper, we propose a Deep Reinforcement learning framework for real-time Event Summarization (DRES), which shows promising performance for resolving all three challenges (i.e., relevance, non-redundancy, timeliness) in a unified framework. Specifically, we (i) devise a hierarchical cross-attention network with intra- and inter-document attentions to integrate important semantic features within and between the query and input document for better text matching. In addition, relevance prediction is leveraged as an auxiliary task to strengthen the document modeling and help to extract relevant documents; (ii) propose a multi-topic dynamic memory network to capture the sequential patterns of different topics belonging to the event of interest and temporally memorize the input facts from the evolving document stream, avoiding extracting redundant information at each time step; (iii) consider both historical dependencies and future uncertainty of the document stream for generating relevant and timely summaries by exploiting the reinforcement learning technique. Experimental results on two real-world datasets have demonstrated the advantages of DRES model with significant improvement in generating relevant, non-redundant, and timely event summaries against the state-of-the-arts.

Emotion-cause pair extraction aims to extract all potential pairs of emotions and corresponding causes from unannotated emotion text.Most existing methods are pipelined framework, which identifies emotions and extracts causes separately, leading to a drawback of error propagation.Towards this issue, we propose a transition-based model to transform the task into a procedure of parsing-like directed graph construction.The proposed model incrementally generates the directed graph with labeled edges based on a sequence of actions, from which we can recognize emotions with the corresponding causes simultaneously, thereby optimizing separate subtasks jointly and maximizing mutual benefits of tasks interdependently.Experimental results show that our approach achieves the best performance, outperforming the state-of-the-art methods by 6.71% (p < 0.01) in F 1 measure.

Stance detection is an important task, which aims to classify the attitude of an opinionated text towards a given target. Remarkable success has been achieved when sufficient labeled training data is available. However, annotating sufficient data is labor-intensive, which establishes significant barriers for generalizing the stance classifier to the data with new targets. In this paper, we proposed a Semantic-Emotion Knowledge Transferring (SEKT) model for cross-target stance detection, which uses the external knowledge (semantic and emotion lexicons) as a bridge to enable knowledge transfer across different targets. Specifically, a semantic-emotion heterogeneous graph is constructed from external semantic and emotion lexicons, which is then fed into a graph convolutional network to learn multi-hop semantic connections between words and emotion tags. Then, the learned semantic-emotion graph representation, which serves as prior knowledge bridging the gap between the source and target domains, is fully integrated into the bidirectional long short-term memory (BiLSTM) stance classifier by adding a novel knowledge-aware memory unit to the BiLSTM cell. Extensive experiments on a large real-world dataset demonstrate the superiority of SEKT against the state-of-the-art baseline methods.

Cross-lingual Machine Reading Comprehension (CLMRC) remains a challenging problem due to the lack of large-scale annotated datasets in low-source languages, such as Arabic, Hindi, and Vietnamese.Many previous approaches use translation data by translating from a rich-source language, such as English, to low-source languages as auxiliary supervision.However, how to effectively leverage translation data and reduce the impact of noise introduced by translation remains onerous.In this paper, we tackle this challenge and enhance the cross-lingual transferring performance by a novel augmentation approach named Language Branch Machine Reading Comprehension (LBMRC).A language branch is a group of passages in one single language paired with questions in all target languages.We train multiple machine reading comprehension (MRC) models proficient in individual language based on LBMRC.Then, we devise a multilingual distillation approach to amalgamate knowledge from multiple language branch models to a single model for all target languages.Combining the LBMRC and multilingual distillation can be more robust to the data noises, therefore, improving the model's cross-lingual ability.Meanwhile, the produced single multilingual model can apply to all target languages, which saves the cost of training, inference, and maintenance for multiple models.Extensive experiments on two CLMRC benchmarks clearly show the effectiveness of our proposed method.

Existing end-to-end task-oriented dialog systems struggle to dynamically model long dialog context for interactions and effectively incorporate knowledge base (KB) information into dialog generation. To conquer these limitations, we propose a Dual Dynamic Memory Network (DDMN) for multi-turn dialog generation, which maintains two core components: dialog memory manager and KB memory manager. The dialog memory manager dynamically expands the dialog memory turn by turn and keeps track of dialog history with an updating mechanism, which encourages the model to filter irrelevant dialog history and memorize important newly coming information. The KB memory manager shares the structural KB triples throughout the whole conversation, and dynamically extracts KB information with a memory pointer at each turn. Experimental results on three benchmark datasets demonstrate that DDMN significantly outperforms the strong baselines in terms of both automatic evaluation and human evaluation. Our code is available at https://github.com/siat-nlp/DDMN.

Image paragraph captioning (IPC) aims to generate a fine-grained paragraph to describe the visual content of an image.Significant progress has been made by deep neural networks, in which the attention mechanism plays an essential role.However, conventional attention mechanisms tend to ignore the past alignment information, which often results in problems of repetitive captioning and incomplete captioning.In this paper, we propose an Interactive key-value Memoryaugmented Attention model for image Paragraph captioning (IMAP) to keep track of the attention history (salient objects coverage information) along with the update-chain of the decoder state and therefore avoid generating repetitive or incomplete image descriptions.In addition, we employ an adaptive attention mechanism to realize adaptive alignment from image regions to caption words, where an image region can be mapped to an arbitrary number of caption words while a caption word can also attend to an arbitrary number of image regions.Extensive experiments on a benchmark dataset (i.e., Stanford) demonstrate the effectiveness of our IMAP model.

The challenge of both achieving task completion by querying the knowledge base and generating human-like responses for task-oriented dialogue systems is attracting increasing research attention.In this paper, we propose a "Two-Teacher One-Student" learning framework (TTOS) for task-oriented dialogue, with the goal of retrieving accurate KB entities and generating human-like responses simultaneously.TTOS amalgamates knowledge from two teacher networks that together provide comprehensive guidance to build a highquality task-oriented dialogue system (student network).Each teacher network is trained via reinforcement learning with a goal-specific reward, which can be viewed as an expert towards the goal and transfers the professional characteristic to the student network.Instead of adopting the classic student-teacher learning of forcing the output of a student network to exactly mimic the soft targets produced by the teacher networks, we introduce two discriminators as in generative adversarial network (GAN) to transfer knowledge from two teachers to the student.The usage of discriminators relaxes the rigid coupling between the student and teachers.Extensive experiments on two benchmark datasets (i.e., CamRest and In-Car Assistant) demonstrate that TTOS significantly outperforms baseline methods.For reproducibility, we release the code and data at https://github.com/siat-nlp/TTOS.

Pre-trained language models (e.g., BERT) have achieved significant success in various natural language processing (NLP) tasks.However, high storage and computational costs obstruct pre-trained language models to be effectively deployed on resourceconstrained devices.In this paper, we propose a novel BERT distillation method based on many-to-many layer mapping, which allows each intermediate student layer to learn from any intermediate teacher layers.In this way, our model can learn from different teacher layers adaptively for various NLP tasks.In addition, we leverage Earth Mover's Distance (EMD) to compute the minimum cumulative cost that must be paid to transform knowledge from teacher network to student network.EMD enables the effective matching for many-to-many layer mapping.Furthermore, we propose a cost attention mechanism to learn the layer weights used in EMD automatically, which is supposed to further improve the model's performance and accelerate convergence time.Extensive experiments on GLUE benchmark demonstrate that our model achieves competitive performance compared to strong competitors in terms of both accuracy and model compression.For reproducibility, we release the code and data at

Question answering (QA) in the legal domain has gained increasing popularity for people to seek legal advice. However, existing QA systems struggle to comprehend the legal context and provide jurisdictionally relevant answers due to the lack of domain expertise. In this paper, we develop an Artificial Intelligence Law Assistant (AILA) for question answering in the domain of Chinese laws. AILA system automatically comprehends users' natural language queries with the help of the legal knowledge graph (KG) and provides the best matching answers for given queries. In addition, AILA provides visual cues to interpret the input queries and candidate answers based on the legal KG. Experimental results on a large-scale legal QA corpus show the effectiveness of AILA. To the best of our knowledge, AILA is the first Chinese legal QA system which integrates the domain knowledge from legal KG to comprehend the questions and answers for ranking QA pairs. AILA is available at http://bmilab.ticp.io:48478/.

In this paper, we propose a multi-source domain adaptation method with a Granger-causal objective (MDA-GC) for cross-domain sentiment classification. Specifically, for each source domain, we build an expert model by using a novel sentiment-guided capsule network, which captures the domain invariant knowledge that bridges the knowledge gap between the source and target domains. Then, an attention mechanism is devised to assign importance weights to a mixture of experts, each of which specializes in a different source domain. In addition, we propose a Granger causal objective to make the weights assigned to individual experts correlate strongly with their contributions to the decision at hand. Experimental results on a benchmark dataset demonstrate that the proposed MDA-GC model significantly outperforms the compared methods.

Non-factoid question answering (QA) is one of the most extensive yet challenging application and research areas of retrieval-based question answering. In particular, answers to non-factoid questions can often be too lengthy and redundant to comprehend, which leads to the great demand on answer sumamrization in non-factoid QA. However, the multi-level interactions between QA pairs and the interrelation among different answer sentences are usually modeled separately on current answer summarization studies. In this paper, we propose a unified model to bridge hierarchical and sequential context modeling for question-driven extractive answer summarization. Specifically, we design a hierarchical compare-aggregate method to integrate the interaction between QA pairs in both word-level and sentence-level into the final question and answer representations. After that, we conduct the question-aware sequential extractor to produce a summary for the lengthy answer. Experimental results show that answer summarization benefits from both hierarchical and sequential context modeling and our method achieves superior performance on WikiHowQA and PubMedQA.

Sequential recommender systems (SRS) have become the key technology in capturing user's dynamic interests and generating high-quality recommendations. Current state-of-the-art sequential recommender models are typically based on a sandwich-structured deep neural network, where one or more middle (hidden) layers are placed between the input embedding layer and output softmax layer. In general, these models require a large number of parameters to obtain optimal performance. Despite the effectiveness, at some point, further increasing model size may be harder for model deployment in resource-constraint devices. To resolve the issues, we propose a compressed sequential recommendation framework, termed as CpRec, where two generic model shrinking techniques are employed. Specifically, we first propose a block-wise adaptive decomposition to approximate the input and softmax matrices by exploiting the fact that items in SRS obey a long-tailed distribution. To reduce the parameters of the middle layers, we introduce three layer-wise parameter sharing schemes. We instantiate CpRec using deep convolutional neural network with dilated kernels given consideration to both recommendation accuracy and efficiency. By the extensive ablation studies, we demonstrate that the proposed CpRec can achieve up to 4~8 times compression rates in real-world SRS datasets. Meanwhile, CpRec is faster during training & inference, and in most cases outperforms its uncompressed counterpart.

Chinese word segmentation (CWS) is an important research topic in information retrieval (IR) and natural language processing (NLP). Significant progresses have been made by deep neural networks with context features. However, these deep models may fail to deal with rare or ambiguous words, thus limit the overall CWS performance. In this paper, we propose a lexicon-enhanced adaptive attention network (LAAN), which takes full advantage of external lexicons to deal with the rare or ambiguous words. Specifically, we devise an adaptive attention mechanism to learn the lexicon-aware representation. In addition, we propose a fusion gate to effectively integrate the additional word information with context information to improve the performance of CWS. LAAN is evaluated on four benchmark datasets, and the experimental results demonstrate that LAAN has robust superiority over the compared methods.

Temporal link prediction in dynamic networks has attracted increasing attention recently due to its valuable real-world applications. The primary challenge of temporal link prediction is to capture the spatial-temporal patterns and high nonlinearity of dynamic networks. Inspired by the success of image generation, we convert the dynamic network into a sequence of static images and formulate the temporal link prediction as a conditional image generation problem. We propose a novel deep generative framework, called NetworkGAN, to tackle the challenging temporal link prediction task efficiently, which simultaneously models the spatial and temporal features in the dynamic networks via deep learning techniques. The proposed NetworkGAN inherits the advantages of the graph convolutional network (GCN), the temporal matrix factorization (TMF), the long short-term memory network (LSTM), and the generative adversarial network (GAN). Specifically, an attentive GCN is first designed to automatically learn the spatial features of dynamic networks. Second, we propose a TMF enhanced attentive LSTM (TMF-LSTM) to capture the temporal dependencies and evolutionary patterns of dynamic networks, which predicts the network snapshot at next timestamp based on the network snapshots observed at previous timestamps. Furthermore, we employ a GAN framework to further refine the performance of temporal link prediction by using a discriminative model to guide the training of the deep generative model (i.e., TMF-LSTM) in an adversarial process. To verify the effectiveness of the proposed model, we conduct extensive experiments on five real-world datasets. Experimental results demonstrate the significant advantages of NetworkGAN compared to other strong competitors.

Movie recommendation systems provide users with ranked lists of movies based on individual's preferences and constraints. Two types of models are commonly used to generate ranking results: 1) long-term models and 2) session-based models. The long-term-based models represent the interactions between users and movies that are supposed to change slowly across time, while the session-based models encode the information of users' interests and changing dynamics of movies' attributes in short terms. In this paper, we propose the LSIC model, leveraging long and short-term information for content-aware movie recommendation using adversarial training. In the adversarial process, we train a generator as an agent of reinforcement learning which recommends the next movie to a user sequentially. We also train a discriminator which attempts to distinguish the generated list of movies from the real records. The poster information of movies is integrated to further improve the performance of movie recommendation, which is specifically essential when few ratings are available. The experiments demonstrate that the proposed model has robust superiority over competitors and achieves the state-of-the-art results.

Image captioning, which aims to generate a sentence to describe the key content of a query image, is an important but challenging task. Existing image captioning approaches can be categorised into two types: generation-based methods and retrieval-based methods. Retrieval-based methods describe images by retrieving pre-existing captions from a repository. Generation-based methods synthesize a new sentence that verbalizes the query image. Both ways have certain advantages but suffer from their own disadvantages. In the paper, we propose a novel EnsCaption model, which aims at enhancing an ensemble of retrieval-based and generation-based image captioning methods through a novel dual generator generative adversarial network. Specifically, EnsCaption is composed of a caption generation model that synthesizes tailored captions for the query image, a caption re-ranking model that retrieves the best-matching caption from a candidate caption pool consisting of generated captions and pre-retrieved captions, and a discriminator that learns the multi-level difference between the generated/retrieved captions and the ground-truth captions. During the adversarial training process, the caption generation model and the caption re-ranking model provide improved synthetic and retrieved candidate captions with high ranking scores from the discriminator, while the discriminator based on multi-level ranking is trained to assign low ranking scores to the generated and retrieved image captions. Our model absorbs the merits of both generation-based and retrieval-based approaches. We conduct comprehensive experiments to evaluate the performance of EnsCaption on two benchmark datasets: MSCOCO and Flickr-30K. Experimental results show that EnsCaption achieves impressive performance compared to the strong baseline methods.

Answer selection and knowledge base question answering (KBQA) are two important tasks of question answering (QA) systems. Existing methods solve these two tasks separately, which requires large number of repetitive work and neglects the rich correlation information between tasks. In this paper, we tackle answer selection and KBQA tasks simultaneously via multi-task learning (MTL), motivated by the following motivations. First, both answer selection and KBQA can be regarded as a ranking problem, with one at text-level while the other at knowledge-level. Second, these two tasks can benefit each other: answer selection can incorporate the external knowledge from knowledge base (KB), while KBQA can be improved by learning contextual information from answer selection. To fulfill the goal of jointly learning these two tasks, we propose a novel multi-task learning scheme that utilizes multi-view attention learned from various perspectives to enable these tasks to interact with each other as well as learn more comprehensive sentence representations. The experiments conducted on several real-world datasets demonstrate the effectiveness of the proposed method, and the performance of answer selection and KBQA is improved. Also, the multi-view attention scheme is proved to be effective in assembling attentive information from different representational perspectives.

In this paper, we propose a Multi-Task learning approach for Answer Selection (MTAS), motivated by the fact that humans have no difficulty performing such task because they possess capabilities of multiple domains (tasks). Specifically, MTAS consists of two key components: (i) A category classification model that learns rich category-aware document representation; (ii) An answer selection model that provides the matching scores of question-answer pairs. These two tasks work on a shared document encoding layer, and they cooperate to learn a high-quality answer selection system. In addition, a multi-head attention mechanism is proposed to learn important information from different representation subspaces at different positions. We manually annotate the first Chinese question answering dataset in law domain (denoted as LawQA) to evaluate the effectiveness of our model. The experimental results show that our model MTAS consistently outperforms the compared methods.1

In this paper, we propose a novel Human-like Semantic Cognition Network (HSCN) for aspect-level sentiment classification, motivated by the principles of human beings’ reading cognitive process (pre-reading, active reading, post-reading). We first design a word-level interactive perception module to capture the correlation between context words and the given target words, which can be regarded as pre-reading. Second, to mimic the process of active reading, we propose a targetaware semantic distillation module to produce the targetspecific context representation for aspect-level sentiment prediction. Third, we further devise a semantic deviation metric module to measure the semantic deviation between the targetspecific context representation and the given target, which evaluates the degree we understand the target-specific context semantics. The measured semantic deviation is then used to fine-tune the above active reading process in a feedback regulation way. To verify the effectiveness of our approach, we conduct extensive experiments on three widely used datasets. The experiments demonstrate that HSCN achieves impressive results compared to other strong competitors.

We propose a Multi-task learning approach for Abstractive Text Summarization (MATS), motivated by the fact that humans have no difficulty performing such task because they have the capabilities of multiple domains. Specifically, MATS consists of three components: (i) a text categorization model that learns rich category-specific text representations using a bi-LSTM encoder; (ii) a syntax labeling model that learns to improve the syntax-aware LSTM decoder; and (iii) an abstractive text summarization model that shares its encoder and decoder with the text categorization and the syntax labeling tasks, respectively. In particular, the abstractive text summarization model enjoys significant benefit from the additional text categorization and syntax knowledge. Our experimental results show that MATS outperforms the competitors.1

In this paper, we propose a Document Embedding Network (DEN) to learn document embeddings in an unsupervised manner. Our model uses the encoder-decoder architecture as its backbone, which tries to reconstruct the input document from an encoded document embedding. Unlike the standard decoder for text reconstruction, we randomly block some words in the input document, and use the incomplete context information and the encoded document embedding to predict the blocked words in the document, inspired by the crossword game. Thus, our decoder can keep the balance between the known and unknown information, and consider both global and partial information when decoding the missing words. We evaluate the learned document embeddings on two tasks: document classification and document retrieval. The experimental results show that our model substantially outperforms the compared methods.1.

The recent artificial intelligence studies have witnessed great interest in abstractive text summarization. Although remarkable progress has been made by deep neural network based methods, generating plausible and high-quality abstractive summaries remains a challenging task. The human-like reading strategy is rarely explored in abstractive text summarization, which however is able to improve the effectiveness of the summarization by considering the process of reading comprehension and logical thinking. Motivated by the humanlike reading strategy that follows a hierarchical routine, we propose a novel Hybrid learning model for Abstractive Text Summarization (HATS). The model consists of three major components, a knowledge-based attention network, a multitask encoder-decoder network, and a generative adversarial network, which are consistent with the different stages of the human-like reading strategy. To verify the effectiveness of HATS, we conduct extensive experiments on two real-life datasets, CNN/Daily Mail and Gigaword datasets. The experimental results demonstrate that HATS achieves impressive results on both datasets.

Obstacles hindering the development of capsule networks for challenging NLP applications include poor scalability to large output spaces and less reliable routing processes.In this paper, we introduce (i) an agreement score to evaluate the performance of routing processes at instance level; (ii) an adaptive optimizer to enhance the reliability of routing; (iii) capsule compression and partial routing to improve the scalability of capsule networks.We validate our approach on two NLP tasks, namely: multi-label text classification and question answering.Experimental results show that our approach considerably improves over strong competitors on both tasks.In addition, we gain the best results in low-resource settings with few training instances.1

Textual entailment is a central problem of language variability, which has been attracting a lot of interest and it poses significant issues in front of systems aimed at natural language understanding. Recently, various frameworks have been proposed for textual entailment recognition, ranging from traditional computational linguistics techniques to deep learning model based methods. However, recent deep neural networks that achieve the state of the art on textual entailment task only consider the context information of the given sentences rather than the real-world background information and knowledge beyond the context. In the paper, we propose a Knowledge-Context Interactive Textual Entailment Network (KCI-TEN) that learns graph level sentence representations by harnessing external knowledge graph with graph attention network. We further propose a text-graph interaction mechanism for neural based entailment matching learning, which endows the redundancy and noise with less importance and put emphasis on the informative representations. Experiments on the SciTail dataset demonstrate that KCI-TEN outperforms the state-of-the-art methods.

Knowledge Graph (KG) contains entities and the relations between entities. Due to its representation ability, KG has been successfully applied to support many medical/healthcare tasks. However, in the medical domain, knowledge holds under certain conditions. Such conditions for medical knowledge are crucial for decision-making in various medical applications, which is missing in existing medical KGs. In this paper, we aim to discovery medical knowledge conditions from texts to enrich KGs. Electronic Medical Records (EMRs) are systematized collection of clinical data and contain detailed information about patients, thus EMRs can be a good resource to discover medical knowledge conditions. Unfortunately, the amount of available EMRs is limited due to reasons such as regularization. Meanwhile, a large amount of medical question answering (QA) data is available, which can greatly help the studied task. However, the quality of medical QA data is quite diverse, which may degrade the quality of the discovered medical knowledge conditions. In the light of these challenges, we propose a new truth discovery method, MedTruth, for medical knowledge condition discovery, which incorporates prior source quality information into the source reliability estimation procedure, and also utilizes the knowledge triple information for trustworthy information computation. We conduct series of experiments on real-world medical datasets to demonstrate that the proposed method can discover meaningful and accurate conditions for medical knowledge by leveraging both EMR and QA data. Further, the proposed method is tested on synthetic datasets to validate its effectiveness under various scenarios.

In this paper, we propose a multi-task learning approach for cross-modal image-text retrieval. First, a correlation network is proposed for relation recognition task, which helps learn the complicated relations and common information of different modalities. Then, we propose a correspondence cross-modal autoencoder for cross-modal input reconstruction task, which helps correlate the hidden representations of two uni-modal autoencoders. In addition, to further improve the performance of cross-modal retrieval, two regularization terms (variance and consistency constraints) are introduced to the cross-modal embeddings such that the learned common information has large variance and is modality invariant. Finally, to enable large-scale cross-modal similarity search, a flexible binary transform network is designed to convert the text and image embeddings into binary codes. Extensive experiments on two benchmark datasets demonstrate that our model has robust superiority over the compared strong baseline methods. Source code is available at \urlhttps://github.com/daerv/DAEVR.

Recent image captioning approaches are typically trained on generation-based or retrieval-based approaches. Both methods have their advantages but limited by the disadvantages. In this paper, we propose a Unified Generation-Retrieval framework for Image Captioning (UGRIC) by using adversarial learning. Different from previous methods, the proposed UGRIC model leverages the informative contents of N-best response candidates provided by the retrieval-based model to enhance the generation-based method. In addition, to further improve the informativeness of the generated caption, we employ copying mechanism to choose words from the retrieved candidate captions and put them into proper positions of the output sequence. Experiments on MSCOCO dataset demonstrate the effectiveness of the UGRIC model through various evaluation metrics.\footnoteCode and data are available at: \urlhttp://tinyurl.com/y6z2x6ho.

Chuang Fan, Hongyu Yan, Jiachen Du, Lin Gui, Lidong Bing, Min Yang, Ruifeng Xu, Ruibin Mao. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Qingnan Jiang, Lei Chen, Ruifeng Xu, Xiang Ao, Min Yang. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Ling Luo, Xiang Ao, Yan Song, Feiyang Pan, Min Yang, Qing He. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

The last decade has witnessed the explosive growth of malicious Internet domains which serve as the fundamental infrastructure for establishing advanced persistent threat command and control communication channels or hosting phishing Web sites. Given the big data nature of Internet traffic data and the ability of algorithmically generating domains and acquiring and registering the domains in a near-automated fashion, detecting malicious domains in real-time is a daunting task for security analysts and network operators. In this paper, we introduce bipartite graphs to capture the interactions between end hosts and domains, identify associated IP addresses of domains, and characterize time-series patterns of DNS queries for domains, and explore one-mode projections of these bipartite graphs for modeling the behavioral, IP-structural, and temporal similarities between domains. We employ graph embedding technique to automatically learn dynamic and discriminative feature representations for over 10,000 labeled domains, and develop an SVM-based classification algorithm for predicting malicious or benign domains. Our model makes the progress towards adapting to the changing and evolving strategies of malicious domains. The experimental results have shown that our proposed algorithm achieves an area under the curve (AUC) of 0.94 based on k-fold cross-validation. To the best of our knowledge, this is the first effort to apply the combination of behavioral modeling and graph embedding for effectively and accurately detecting malicious domains.

In this paper, we propose a Knowledge-enhanced Hierarchical Attention for community question answering with Multi-task learning and Adaptive learning (KHAMA). First, we propose a hierarchical attention network to fully fuse knowledge from input documents and knowledge base (KB) by exploiting the semantic compositionality of the input sequences. The external factual knowledge helps recognize background knowledge (entity mentions and their relationships) and eliminate noise information from long documents that have sophisticated syntactic and semantic structures. In addition, we build multiple CQA models with adaptive boosting and then combine these models to learn a more effective and robust CQA system. Further- more, KHAMA is a multi-task learning model. It regards CQA as the primary task and question categorization as the auxiliary task, aiming at learning a category-aware document encoder and enhance the quality of identifying essential information from long questions. Extensive experiments on two benchmarks demonstrate that KHAMA achieves substantial improvements over the compared methods.

In this paper, we generally formulate the dynamics prediction problem of various network systems (e.g., the prediction of mobility, traffic and topology) as the temporal link prediction task. Different from conventional techniques of temporal link prediction that ignore the potential non-linear characteristics and the informative link weights in the dynamic network, we introduce a novel non-linear model GCN-GAN to tackle the challenging temporal link prediction task of weighted dynamic networks. The proposed model leverages the benefits of the graph convolutional network (GCN), long short-term memory (LSTM) as well as the generative adversarial network (GAN). Thus, the dynamics, topology structure and evolutionary patterns of weighted dynamic networks can be fully exploited to improve the temporal link prediction performance. Concretely, we first utilize GCN to explore the local topological characteristics of each single snapshot and then employ LSTM to characterize the evolving features of the dynamic networks. Moreover, GAN is used to enhance the ability of the model to generate the next weighted network snapshot, which can effectively tackle the sparsity and the wide-value-range problem of edge weights in real-life dynamic networks. To verify the model's effectiveness, we conduct extensive experiments on four datasets of different network systems and application scenarios. The experimental results demonstrate that our model achieves impressive results compared to the state-of-the-art competitors.

Knowledge graphs (KGs) have been applied to many semantic-driven applications, including knowledge interchange and semantic inference. However, most KGs are far from complete and are growing rapidly. Although significant progress has been made in the symbolic representation learning of KGs with structural information, the textual knowledge that plays a crucial role in relation prediction is underutilized, and the issues of redundancy and noise path remain to be settled. In this paper, a Path-based Attribute-aware Representation Learning model (PARL) has been proposed to perform path denoising and path representation learning for the relation prediction task. We develop a novel text-enhanced relation prediction architecture, which interactively learns KG structural and textual representations to vary the sparsity and reliability of KG. Moreover, a path denoising algorithm is presented to emphasize paths with rich information and reduce the impact of redundancy and noise path. Experiments on a public dataset demonstrate that PARL consistently outperforms state-of-the-art methods on relation prediction and KG completion tasks.

Knowledge Based Question Answering (KBQA) is one of the most promising approaches to provide suitable answers for the queries posted by users. Relation detection that aims to take full advantage of the substantial knowledge contained in knowledge base (KB) becomes increasingly important. Significant progress has been made in performing relation detection over KB. However, recent deep neural networks that achieve the state of the art on KB-based relation detection task only consider the context information of question sentences rather than the relatedness between question and answer candidates, and exclusively extract the relation from KB triple rather than learn informative relational path. In this paper, we propose a Knowledge-driven Relation Detection network (KRD) to interactively learn answer-enhanced question representations and path-aware relation representations for relation detection. A Siamese LSTM is employed into a similarity matching process between the question representation and relation representation. Experimental results on the SimpleQuestions and WebQSP datasets demonstrate that KRD outperforms the state-of-the-art methods. In addition, a series of ablation test show the robust superiority of the proposed method.

In this paper, we develop a neural attentive interpretable recommendation system, named NAIRS. A self-attention network, as a key component of the system, is designed to assign attention weights to interacted items of a user. This attention mechanism can distinguish the importance of the various interacted items in contributing to a user profile. %, and it also provides interpretable recommendations. Based on the user profiles obtained by the self-attention network, NAIRS offers personalized high-quality recommendation. Moreover, it develops visual cues to interpret recommendations. This demo application with the implementation of NAIRS enables users to interact with a recommendation system, and it persistently collects training data to improve the system. The demonstration and experimental results show the effectiveness of NAIRS.

Abstractive meeting summarization is a challenging problem in natural language understanding, which automatically generates the condensed summary covering the important points in the meeting conversation. However, the existing abstractive summarization works mainly focus on the structured text documents, which may be ineffectively applied to the meeting summarization task due to the lack of modeling the unstructured long-form conversational contents. In this paper, we consider the problem of abstractive meeting summarization from the viewpoint of hierarchical adaptive segmental encoder-decoder network learning. We propose the hierarchical neural encoder based on adaptive recurrent networks to learn the semantic representation of meeting conversation with adaptive conversation segmentation. We then develop the reinforced decoder network to generate the high-quality summaries for abstractive meeting summarization. We conduct the extensive experiments on the well-known AMI meeting conversation dataset to validate the effectiveness of our proposed method.

BACKGROUND: Ontology has attracted substantial attention from both academia and industry. Handling uncertainty reasoning is important in researching ontology. For example, when a patient is suffering from cirrhosis, the appearance of abdominal vein varices is four times more likely than the presence of bitter taste. Such medical knowledge is crucial for decision-making in various medical applications but is missing from existing medical ontologies. In this paper, we aim to discover medical knowledge probabilities from electronic medical record (EMR) texts to enrich ontologies. First, we build an ontology by identifying meaningful entity mentions from EMRs. Then, we propose a symptom-dependency-aware naïve Bayes classifier (SDNB) that is based on the assumption that there is a level of dependency among symptoms. To ensure the accuracy of the diagnostic classification, we incorporate the probability of a disease into the ontology via innovative approaches. RESULTS: We conduct a series of experiments to evaluate whether the proposed method can discover meaningful and accurate probabilities for medical knowledge. Based on over 30,000 deidentified medical records, we explore 336 abdominal diseases and 81 related symptoms. Among these 336 gastrointestinal diseases, the probabilities of 31 diseases are obtained via our method. These 31 probabilities of diseases and 189 conditional probabilities between diseases and the symptoms are added into the generated ontology. CONCLUSION: In this paper, we propose a medical knowledge probability discovery method that is based on the analysis and extraction of EMR text data for enriching a medical ontology with probability information. The experimental results demonstrate that the proposed method can effectively identify accurate medical knowledge probability information from EMR data. In addition, the proposed method can efficiently and accurately calculate the probability of a patient suffering from a specified disease, thereby demonstrating the advantage of combining an ontology and a symptom-dependency-aware naïve Bayes classifier.

Abstract Unlike English and other Western languages, many Asian languages such as Chinese and Japanese do not delimit words by space. Word segmentation and new word detection are therefore key steps in processing these languages. Chinese word segmentation can be considered as a part-of-speech (POS)-tagging problem. We can segment corpus by assigning a label for each character which indicates the position of the character in a word (e.g., “B” for word beginning, and “E” for the end of the word, etc.). Chinese word segmentation seems to be well studied. Machine learning models such as conditional random field (CRF) and bi-directional long short-term memory (LSTM) have shown outstanding performances on this task. However, the segmentation accuracies drop significantly when applying the same approaches to out-domain cases, in which high-quality in-domain training data are not available. An example of out-domain applications is the new word detection in Chinese microblogs for which the availability of high-quality corpus is limited. In this paper, we focus on out-domain Chinese new word detection. We first design a new method Edge Likelihood (EL) for Chinese word boundary detection. Then we propose a domain-independent Chinese new word detector (DICND); each Chinese character is represented as a low-dimensional vector in the proposed framework, and segmentation-related features of the character are used as the values in the vector.

Microarray technology enables the collection of vast amounts of gene expression data from biological experiments. Clustering algorithms have been successfully applied to exploring the gene expression data. Since a set of genes may be only correlated to a subset of samples, it is useful to use co-clustering to recover co-clusters in the gene expression data. In this paper, we propose a novel algorithm, called Subspace Weighting Co-Clustering (SWCC), for high dimensional gene expression data. In SWCC, a gene subspace weight matrix is introduced to identify the contribution of gene objects in distinguishing different sample clusters. We design a new co-clustering objective function to recover the co-clusters in the gene expression data, in which the subspace weight matrix is introduced. An iterative algorithm is developed to solve the objective function, in which the subspace weight matrix is automatically computed during the iterative co-clustering process. Our empirical study shows encouraging results of the proposed algorithm in comparison with six state-of-the-art clustering algorithms on ten gene expression data sets. We also propose to use SWCC for gene clustering and selection. The experimental results show that the selected genes can improve the classification performance of Random Forests.

Recent artificial intelligence research has witnessed great interest in automatically generating text descriptions of images, which are known as the image captioning task. Remarkable success has been achieved on domains where a large number of paired data in multimedia are available. Nevertheless, annotating sufficient data is labor-intensive and time-consuming, establishing significant barriers for adapting the image captioning systems to new domains. In this study, we introduc a novel Multitask Learning Algorithm for cross-Domain Image Captioning (MLADIC). MLADIC is a multitask system that simultaneously optimizes two coupled objectives via a dual learning mechanism: image captioning and text-to-image synthesis, with the hope that by leveraging the correlation of the two dual tasks, we are able to enhance the image captioning performance in the target domain. Concretely, the image captioning task is trained with an encoder-decoder model (i.e., CNN-LSTM) to generate textual descriptions of the input images. The image synthesis task employs the conditional generative adversarial network (C-GAN) to synthesize plausible images based on text descriptions. In C-GAN, a generative model $G$ synthesizes plausible images given text descriptions, and a discriminative model $D$ tries to distinguish the images in training data from the generated images by $G$. The adversarial process can eventually guide $G$ to generate plausible and high-quality images. To bridge the gap between different domains, a two-step strategy is adopted in order to transfer knowledge from the source domains to the target domains. First, we pre-train the model to learn the alignment between the neural representations of images and that of text data with the sufficient labeled source domain data. Second, we fine-tune the learned model by leveraging the limited image-text pairs and unpaired data in the target domain. We conduct extensive experiments to evaluate the performance of MLADIC by using the MSCOCO as the source domain data, and using Flickr30k and Oxford-102 as the target domain data. The results demonstrate that MLADIC achieves substantially better performance than the strong competitors for the cross-domain image captioning task.

Deep neural networks have gained great success recently for sentiment classification. However, these approaches do not fully exploit the linguistic knowledge. In this paper, we propose a novel sentiment lexicon enhanced attention-based LSTM (SLEA-LSTM) model to improve the performance of sentence-level sentiment classification. Our method successfully integrates sentiment lexicon into deep neural networks via single-head or multi-head attention mechanisms. We conduct extensive experiments on MR and SST datasets. The experimental results show that our model achieved comparable or better performance than the state-of-the-art methods.

In this paper, we propose an adversarial process for abstractive text summarization, in which we simultaneously train a generative model G and a discriminative model D. In particular, we build the generator G as an agent of reinforcement learning, which takes the raw text as input and predicts the abstractive summarization. We also build a discriminator which attempts to distinguish the generated summary from the ground truth summary. Extensive experiments demonstrate that our model achieves competitive ROUGE scores with the state-of-the-art methods on CNN/Daily Mail dataset. Qualitatively, we show that our model is able to generate more abstractive, readable and diverse summaries.

Protein complex is a group of associated polypeptide chains which plays essential roles in biological process. Given a graph representing protein-protein interactions (PPI) network, it is critical but non-trivial to detect protein complexes.In this paper, we propose a semi-supervised network embedding model by adopting graph convolutional networks to effectively detect densely connected subgraphs. We conduct extensive experiment on two popular PPI networks with various data sizes and densities. The experimental results show our approach achieves state-of-the-art performance.

This study takes the lead to study the aspect/sentiment-aware abstractive review summarization in domain adaptation scenario. The proposed model CASAS (neural attentive model for Cross-domain Aspect/Sentiment-aware Abstractive review Summarization) leverages domain classification task, working on datasets of both source and target domains, to recognize the domain information of texts and transfer knowledge from source domains to target domains. The extensive experiments on Amazon reviews demonstrate that CASAS outperforms the compared methods in both out-of-domain and in-domain setups.

Answer selection is an important but challenging task. Significant progresses have been made in domains where a large amount of labeled training data is available. However, obtaining rich annotated data is a time-consuming and expensive process, creating a substantial barrier for applying answer selection models to a new domain which has limited labeled data. In this paper, we propose Knowledge-aware Attentive Network (KAN), a transfer learning framework for cross-domain answer selection, which uses the knowledge base as a bridge to enable knowledge transfer from the source domain to the target domains. Specifically, we design a knowledge module to integrate the knowledge-based representational learning into answer selection models. The learned knowledge-based representations are shared by source and target domains, which not only leverages large amounts of cross-domain data, but also benefits from a regularization effect that leads to more general representations to help tasks in new domains. To verify the effectiveness of our model, we use SQuAD-T dataset as the source domain and three other datasets (i.e., Yahoo QA, TREC QA and InsuranceQA) as the target domains. The experimental results demonstrate that KAN has remarkable applicability and generality, and consistently outperforms the strong competitors by a noticeable margin for cross-domain answer selection.

Distantly supervised relation extraction greatly reduces human efforts in extracting relational facts from unstructured texts. However, it suffers from noisy labeling problem, which can degrade its performance. Meanwhile, the useful information expressed in knowledge graph is still underutilized in the state-of-the-art methods for distantly supervised relation extraction. In the light of these challenges, we propose CORD, a novelCOopeRativeDenoising framework, which consists two base networks leveraging text corpus and knowledge graph respectively, and a cooperative module involving their mutual learning by the adaptive bi-directional knowledge distillation and dynamic ensemble with noisy-varying instances. Experimental results on a real-world dataset demonstrate that the proposed method reduces the noisy labels and achieves substantial improvement over the state-of-the-art methods.

Review text has been widely studied in traditional tasks such as sentiment analysis and aspect extraction. However, to date, no work is towards the abstractive review summarization that is essential for business organizations and individual consumers to make informed decisions. This work takes the lead to study the aspect/sentiment-aware abstractive review summarization by exploring multi-factor attentions. Specifically, we propose an interactive attention mechanism to interactively learns the representations of context words, sentiment words and aspect words within the reviews, acted as an encoder. The learned sentiment and aspect representations are incorporated into the decoder to generate aspect/sentiment-aware review summaries via an attention fusion network. In addition, the abstractive summarizer is jointly trained with the text categorization task, which helps learn a category-specific text encoder, locating salient aspect information and exploring the variations of style and wording of content with respect to different text categories. The experimental results on a real-life dataset demonstrate that our model achieves impressive results compared to other strong competitors.

In this study, we explore capsule networks with dynamic routing for text classification.We propose three strategies to stabilize the dynamic routing process to alleviate the disturbance of some noise capsules which may contain "background" information or have not been successfully trained.A series of experiments are conducted with capsule networks on six text classification benchmarks.Capsule networks achieve competitive results over the compared baseline methods on 4 out of 6 datasets, which shows the effectiveness of capsule networks for text classification.We additionally show that capsule networks exhibit significant improvement when transfer single-label to multi-label text classification over the competitors.To the best of our knowledge, this is the first work that capsule networks have been empirically investigated for text modeling 1 .

Retweet prediction is a challenging problem in social media sites (SMS). In this paper, we study the problem of image retweet prediction in social media, which predicts the image sharing behavior that the user reposts the image tweets from their followees. Unlike previous studies, we learn user preference ranking model from their past retweeted image tweets in SMS. We first propose heterogeneous image retweet modeling network (IRM) that exploits users' past retweeted image tweets with associated contexts, their following relations in SMS and preference of their followees. We then develop a novel attentional multi-faceted ranking network learning framework with multi-modal neural networks for the proposed heterogenous IRM network to learn the joint image tweet representations and user preference representations for prediction task. The extensive experiments on a large-scale dataset from Twitter site shows that our method achieves better performance than other state-of-the-art solutions to the problem.

Recommender systems provide users with ranked lists of items based on individual's preferences and constraints. Two types of models are commonly used to generate ranking results: long-term models and session-based models. While long-term models represent the interactions between users and items that are supposed to change slowly across time, session-based models encode the information of users' interests and changing dynamics of items' attributes in short terms. In this paper, we propose a PLASTIC model, Prioritizing Long And Short-Term Information in top-n reCommendation using adversarial training. In the adversarial process, we train a generator as an agent of reinforcement learning which recommends the next item to a user sequentially. We also train a discriminator which attempts to distinguish the generated list of items from the real list recorded. Extensive experiments show that our model exhibits significantly better performances on two widely used real-world datasets.

In this paper, we propose a Multi-task Learning Approach for Image Captioning (MLAIC ), motivated by the fact that humans have no difficulty performing such task because they possess capabilities of multiple domains. Specifically, MLAIC consists of three key components: (i) A multi-object classification model that learns rich category-aware image representations using a CNN image encoder; (ii) A syntax generation model that learns better syntax-aware LSTM based decoder; (iii) An image captioning model that generates image descriptions in text, sharing its CNN encoder and LSTM decoder with the object classification task and the syntax generation task, respectively. In particular, the image captioning model can benefit from the additional object categorization and syntax knowledge. To verify the effectiveness of our approach, we conduct extensive experiments on MS-COCO dataset. The experimental results demonstrate that our model achieves impressive results compared to other strong competitors.

Code summarization provides a high level natural language description of the function performed by code, as it can benefit the software maintenance, code categorization and retrieval. To the best of our knowledge, most state-of-the-art approaches follow an encoder-decoder framework which encodes the code into a hidden space and then decode it into natural language space, suffering from two major drawbacks: a) Their encoders only consider the sequential content of code, ignoring the tree structure which is also critical for the task of code summarization; b) Their decoders are typically trained to predict the next word by maximizing the likelihood of next ground-truth word with previous ground-truth word given. However, it is expected to generate the entire sequence from scratch at test time. This discrepancy can cause an exposure bias issue, making the learnt decoder suboptimal. In this paper, we incorporate an abstract syntax tree structure as well as sequential content of code snippets into a deep reinforcement learning framework (i.e., actor-critic network). The actor network provides the confidence of predicting the next word according to current state. On the other hand, the critic network evaluates the reward value of all possible extensions of the current state and can provide global guidance for explorations. We employ an advantage reward composed of BLEU metric to train both networks. Comprehensive experiments on a real-world dataset show the effectiveness of our proposed model when compared with some state-of-the-art methods.

During the past decades, many spectral clustering algorithms have been proposed. However, their high computational complexities hinder their applications on large-scale data. Moreover, most of them use a two-step approach to obtain the optimal solution, which may deviate from the solution by directly solving the original problem. In this paper, we propose a new optimization algorithm, namely Direct Normalized Cut (DNC), to directly optimize the normalized cut model. DNC has a quadratic time complexity, which is a significant reduction comparing with the cubic time complexity of the traditional spectral clustering. To cope with large-scale data, a Fast Normalized Cut (FNC) method with linear time and space complexities is proposed by extending DNC with an anchor-based strategy. In the new method, we first seek a set of anchors and then construct a representative similarity matrix by computing distances between the anchors and the whole data set. To find high quality anchors that best represent the whole data set, we propose a Balanced k-means (BKM) to partition a data set into balanced clusters and use the cluster centers as anchors. Then DNC is used to obtain the final clustering result from the representative similarity matrix. A series of experiments were conducted on both synthetic data and real-world data sets, and the experimental results show the superior performance of BKM, DNC and FNC.

In this paper, we propose a personalized dialogue generation system, which combines reinforcement learning techniques with an attention-based hierarchical recurrent encoderdecoder model. Firstly, we incorporate user-specific information into the decoder to capture user's background information and speaking style. Secondly, we employ reinforcement learning techniques to maximize future reward in dialogue, which enables our system to generate topic-coherent, informative and grammatical responses. Moreover, we propose three types of rewards to characterize good conversations. Finally, we compare the performance of the following reinforcement learning methods in dialogue generation: policy gradient, Q-learning, and actor-critic algorithms. We conduct experiments to verify the effectiveness of the proposed model on two dialogue datasets. Experimental results demonstrate that our model can generate better personalized dialogues for different users. Quantitatively, our method achieves better performance than the state-of-the-art dialogue systems in terms of BLEU score, perplexity, and human evaluation.

Analyzing public opinions towards products, services and social events is an important but challenging task. Despite the remarkable successes of deep neural networks in sentiment analysis, these approaches do not make full use of the prior sentiment knowledge (e.g., sentiment lexicon, negation words, intensity words). In this paper, we propose a Sentiment-Aware Attention Network (SAAN) to boost the performance of sentiment analysis, which adopts a three-step strategy to learn the sentiment-specific sentence representation. First, we employ a word-level mutual attention mechanism to model word-level correlation. Next, a phrase-level convolutional attention is designed to obtain phrase-level correlation. Finally, a sentence-level multi-head attention mechanism is proposed to capture various sentimental information from different subspaces. The experiments on Movie Review (MR) and Stanford Sentiment Treebank (SST) show that our model consistently outperform the previous methods for sentiment analysis.

With the rising importance of knowledge exchange, ontologies have become a key technology in the development of shared knowledge models for semantic-driven applications, such as knowledge interchange and semantic integration. Significant progress has been made in the use of entropy to measure the predictability and redundancy of knowledge bases, particularly ontologies. However, the current entropy applications used to evaluate ontologies consider only single-point connectivity rather than path connectivity, assign equal weights to each entity and path, and assume that vertices are static. To address these deficiencies, the present study proposes a Path-based Text-aware Entropy Computation method, PTEC, by considering the path information between different vertices and the textual information within the path to calculate the connectivity path of the whole network and the different weights between various nodes. Information obtained from structure-based embedding and text-based embedding is multiplied by the connectivity matrix of the entropy computation. An experimental evaluation of three real-world ontologies is performed based on ontology statistical information (data quantity), entropy evaluation (data quality), and a case study (ontology structure and text visualization). These aspects mutually demonstrate the reliability of our method. Experimental results demonstrate that PTEC can effectively evaluate ontologies, particularly those in the medical field.

Ranking question answer pairs has attracted increasing attention recently due to its broad applications such as information retrieval and question answering (QA). Significant progresses have been made by deep neural networks. However, background information and hidden relations beyond the context, which play crucial roles in human text comprehension, have received little attention in recent deep neural networks that achieve the state of the art in ranking QA pairs. In the paper, we propose KABLSTM, a Knowledge-aware Attentive Bidirectional Long Short-Term Memory, which leverages external knowledge from knowledge graphs (KG) to enrich the representational learning of QA sentences. Specifically, we develop a context-knowledge interactive learning architecture, in which a context-guided attentive convolutional neural network (CNN) is designed to integrate knowledge embeddings into sentence representations. Besides, a knowledge-aware attention mechanism is presented to attend interrelations between each segments of QA pairs. KABLSTM is evaluated on two widely-used benchmark QA datasets: WikiQA and TREC QA. Experiment results demonstrate that KABLSTM has robust superiority over competitors and sets state-of-the-art.

Clustering of customer transaction data is an important procedure to analyze customer behaviors in retail and e-commerce companies. Note that products from companies are often organized as a product tree, in which the leaf nodes are goods to sell, and the internal nodes (except root node) could be multiple product categories. Based on this tree, we propose the “personalized product tree”, named purchase tree, to represent a customer's transaction records. So the customers' transaction data set can be compressed into a set of purchase trees. We propose a partitional clustering algorithm, named PurTreeClust, for fast clustering of purchase trees. A new distance metric is proposed to effectively compute the distance between two purchase trees. To cluster the purchase tree data, we first rank the purchase trees as candidate representative trees with a novel separate density, and then select the top k customers as the representatives of k customer groups. Finally, the clustering results are obtained by assigning each customer to the nearest representative. We also propose a gap statistic based method to evaluate the number of clusters. A series of experiments were conducted on ten real-life transaction data sets, and experimental results show the superior performance of the proposed method.

Recommending social activities , such as watching movies or having dinner, is a common function found in social networks or e-commerce sites. Besides certain websites which manage activity-related locations (e.g., foursquare.com), many items on product sale platforms (e.g., groupon.com) can naturally be mapped to social activities. For example, movie tickets can be thought of as activity items , which can be mapped as a social activity of “watch a movie.” Traditional recommender systems estimate the degree of interest for a target user on candidate items (or activities), and accordingly, recommend the top- k activity items to the user. However, these systems ignore an important social characteristic of recommended activities: people usually tend to participate in those activities with friends. This article considers this fact for improving the effectiveness of recommendation in two directions. First, we study the problem of activity-partner recommendation ; i.e., for each recommended activity item, find a suitable partner for the user. This (i) saves the user’s time for finding activity partners, (ii) increases the likelihood that the activity item will be selected by the user, and (iii) improves the effectiveness of recommender systems to users overall and enkindles their social enthusiasm. Our partner recommender is built upon the users’ historical attendance preferences, their social context, and geographic information. Moreover, we explore how to leverage the partner recommendation to help improve the effectiveness of recommending activities to users. Assuming that users tend to select the activities for which they can find suitable partners, we propose a partner-aware activity recommendation model, which integrates this hypothesis into conventional recommendation approaches. Finally, the recommended items not only match users’ interests, but also have high chances to be selected by the users, because the users can find suitable partners to attend the corresponding activities together. We conduct experiments on real data to evaluate the effectiveness of activity-partner recommendation and partner-aware activity recommendation. The results verify that (i) suggesting partners greatly improves the likelihood that a recommended activity item is to be selected by the target user and (ii) considering the existence of suitable partners in the ranking of recommended items improves the accuracy of recommendation significantly.

With an increasing number of paid writers posting fake reviews to promote or demote some target entities through Internet, review spammer detection has become a crucial and challenging task. In this paper, we propose a three-phase method to address the problem of identifying review spammer groups and individual spammers, who get paid for posting fake comments. We evaluate the effectiveness and performance of the approach on a real-life online shopping review dataset from amazon.com. The experimental result shows that our model achieved comparable or better performance than previous work on spammer detection.

We present an attention-based bidirectional LSTM approach to improve the target-dependent sentiment classification. Our method learns the alignment between the target entities and the most distinguishing features. We conduct extensive experiments on a real-life dataset. The experimental results show that our model achieves state-of-the-art results.

Authorship attribution is an active research direction due to its legal and financial importance. The goal is to identify the authorship of anonymous texts. In this paper, we propose a Topic Drift Model (TDM), monitoring the dynamicity of authors’ writing style and latent topics of interest. Our model is sensitive to the temporal information and the ordering of words, thus it extracts more information from texts.

Machine Reading and Comprehension recently has drawn a fair amount of attention in the field of natural language processing. In this paper, we consider integrating side information to improve machine comprehension on answering cloze-style questions more precisely. To leverage the external information, we present a novel attention-based architecture which could feed the side information representations into word level embeddings to explore the comprehension performance. Our experiments show consistent improvements of our model over various baselines.

Recent AI research has witnessed increasing interests in automatically generating image descriptions in text, which is coined as theimage captioning problem. Significant progresses have been made in domains where plenty of labeled training data (i.e. image-text pairs) are readily available or collected. However, obtaining rich annotated data is a time-consuming and expensive process, creating a substantial barrier for applying image captioning methods to a new domain. In this paper, we propose a cross-domain image captioning approach that uses a novel dual learning mechanism to overcome this barrier. First, we model the alignment between the neural representations of images and that of natural languages in the source domain where one can access sufficient labeled data. Second, we adjust the pre-trained model based on examining limited data (or unpaired data) in the target domain. In particular, we introduce a dual learning mechanism with a policy gradient method that generates highly rewarded captions. The mechanism simultaneously optimizes two coupled objectives: generating image descriptions in text and generating plausible images from text descriptions, with the hope that by explicitly exploiting their coupled relation, one can safeguard the performance of image captioning in the target domain. To verify the effectiveness of our model, we use MSCOCO dataset as the source domain and two other datasets (Oxford-102 and Flickr30k) as the target domains. The experimental results show that our model consistently outperforms previous methods for cross-domain image captioning.

We study the problem of identifying the topics and sentiments and tracking their shifts from social media texts in different geographical regions during emergencies and disasters.We propose a location-based dynamic sentiment-topic model (LDST) which can jointly model topic, sentiment, time and Geolocation information.The experimental results demonstrate that LDST performs very well at discovering topics and sentiments from social media and tracking their shifts in different geographical regions during emergencies and disasters 1 .

Many spectral clustering algorithms have been proposed and successfully applied to many high-dimensional applications. However, there are still two problems that need to be solved: 1) existing methods for obtaining the final clustering assignments may deviate from the true discrete solution, and 2) most of these methods usually have very high computational complexity. In this paper, we propose a Scalable Normalized Cut method for clustering of large scale data. In the new method, an efficient method is used to construct a small representation matrix and then clustering is performed on the representation matrix. In the clustering process, an improved spectral rotation method is proposed to obtain the solution of the final clustering assignments. A series of experimental were conducted on 14 benchmark data sets and the experimental results show the superior performance of the new method.

In this paper, we propose a novel personalized response generation model via domain adaptation (PRG-DM). First, we learn the human responding style from large general data (without user-specific information). Second, we fine tune the model on a small size of personalized data to generate personalized responses with a dual learning mechanism. Moreover, we propose three new rewards to characterize good conversations that are personalized, informative and grammatical. We employ the policy gradient method to generate highly rewarded responses. Experimental results show that our model can generate better personalized responses for different users.

With the rapid development of mobile devices, point-of-interest (POI) suggestion has become a popular online web service, which provides attractive and interesting locations to users. In order to provide interesting POIs, many existing POI recommendation works learn the latent representations of users and POIs from users' past visiting POIs, which suffers from the sparsity problem of POI data. In this paper, we consider the problem of POI suggestion from the viewpoint of learning geosocial multimedia network representations. We propose a novel max-margin metric geosocial multimedia network representation learning framework by exploiting users' check-in behavior and their social relations. We then develop a random-walk based learning method with max-margin metric network embedding. We evaluate the performance of our method on a large-scale geosocial multimedia network dataset and show that our method achieves the best performance than other state-of-the-art solutions.

Investor social media, such as StockTwist, are gaining increasing popularity. These sites allow users to post their investing opinions and suggestions in the form of microblogs. Given the growth of the posted data, a significant and challenging research problem is how to utilize the personal wisdom and different viewpoints in these opinions to help investment. Previous work aggregates sentiments related to stocks and generates buy or hold recommendations for stocks obtaining favorable votes while suggesting sell or short actions for stocks with negative votes. However, considering the fact that there always exist unreasonable or misleading posts, sentiment aggregation should be improved to be robust to noise. In this paper, we improve investment recommendation by modeling and using the quality of each investment opinion. To model the quality of an opinion, we use multiple categories of features generated from the author information, opinion content and the characteristics of stocks to which the opinion refers. Then, we discuss how to perform investment recommendation (including opinion recommendation and portfolio recommendation) with predicted qualities of investor opinions. Experimental results on real datasets demonstrate effectiveness of our work in recommending high-quality opinions and generating profitable investment decisions.

Discovering the author's interest over time from documents has important applications in recommendation systems, authorship identification and opinion extraction. In this paper, we propose an interest drift model (IDM), which monitors the evolution of author interests in time-stamped documents. The model further uses the discovered author interest information to help finding better topics. Unlike traditional topic models, our model is sensitive to the ordering of words, thus it extracts more information from the semantic meaning of the context. The experiment results show that the IDM model learns better topics than state-of-the-art topic models.

Topic modeling of textual corpora is an important and challenging problem. In most previous work, the “bag-of-words” assumption is usually made which ignores the ordering of words. This assumption simplifies the computation, but it unrealistically loses the ordering information and the semantic of words in the context. In this paper, we present a Gaussian Mixture Neural Topic Model (GMNTM) which incorporates both the ordering of words and the semantic meaning of sentences into topic modeling. Specifically, we represent each topic as a cluster of multi-dimensional vectors and embed the corpus into a collection of vectors generated by the Gaussian mixture model. Each word is affected not only by its topic, but also by the embedding vector of its surrounding words and the context. The Gaussian mixture components and the topic of documents, sentences and words can be learnt jointly. Extensive experiments show that our model can learn better topics and more accurate word distributions for each topic. Quantitatively, comparing to state-of-the-art topic modeling approaches, GMNTM obtains significantly better performance in terms of perplexity, retrieval accuracy and classification accuracy.

Min Yang, Wenting Tu, Ziyu Lu, Wenpeng Yin, Kam-Pui Chow. Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2015.

Analyzing and understanding people's sentiments towards different topics has become an interesting task due to the explosion of opinion-rich resources. In most sentiment analysis applications, sentiment lexicons play a crucial role, to be used as metadata of sentiment polarity. However, most previous works focus on discovering general-purpose sentiment lexicons. They cannot capture domain-specific sentiment words, or implicit and connotative sentiment words that are seemingly objective. In this paper, we propose a supervised sentiment-aware LDA model (ssLDA). The model uses a minimal set of domain-independent seed words and document labels to discover a domain-specific lexicon, learning a lexicon much richer and adaptive to the sentiment of specific document. Experiments on two publicly-available datasets (movie reviews and Obama-McCain debate dataset) show that our model is effective in constructing a comprehensive and high-quality domain-specific sentiment lexicon. Furthermore, the resulting lexicon significantly improves the performance of sentiment classification tasks.