Large Reasoning Models (LRMs) have recently demonstrated impressive performance across a range of reasoning tasks by generating intermediate thoughts. However, these models can suffer from overthinking—generating excessive tokens that contribute little to final accuracy while increasing inference cost. To mitigate this, we propose TIV (Thought Injection via Vectors), an innovative framework that compresses token-level reasoning into compact vectors without sacrificing performance. Rather than generating explicit thoughts, TIV injects learnable vectors into the post-attention hidden states of the final token across Transformer layers, enabling implicit and lightweight reasoning. We further introduce a two-stage reinforcement learning strategy: the first stage calibrates the model's reasoning distribution, and the second distills it into a vector-based policy optimized for both accuracy and brevity. Experiments on three reasoning benchmarks show that TIV preserves over 99% of the original accuracy while reducing output length by more than 65% on average, reaching up to 80% in some cases. Moreover, TIV consistently achieves superior trade-offs between accuracy and efficiency compared to existing methods, distinguishing itself as a state-of-the-art (SOTA) approach for efficient reasoning in LRMs.

Multimodal fake news detection plays a crucial role in combating online misinformation. The inherent domain diversity of news in the real world has driven the development of cross-domain detection methods. However, these detection methods either suffer from significant performance degradation due to semantic and deception pattern shifts between the training (source) and test (target) domains or heavily rely on annotated labels. To address the problems, we propose ADOSE, an active multi-source domain adaptation framework for multimodal fake news detection which actively annotates a small subset of target samples to improve detection performance. Specifically, for domain shifts, we design a multi-expert classifier network based on refined features to comprehensively capture and adapt to the semantic space and deception patterns of news across different domains. To maximize adaptation performance with limited annotation cost, we propose a least-disagree uncertainty selector equipped with a diversity calculator for selecting the most informative samples. The selector leverages the uncertainty of inconsistent predictions before and after perturbations by multiple classifiers as an indicator of unfamiliar samples. It further incorporates diversity scores derived from multi-view features to ensure the chosen samples achieve maximal coverage of target domain features. The extensive experiments on multiple datasets show that ADOSE outperforms existing domain adaptation methods by 2.45% ~ 9.1%, indicating the superiority of our model.

Multimodal sequential recommender systems leverage diverse modal inputs to enhance the accuracy and relevance of personalized recommendations. However, existing fusion strategies often struggle to capture intricate cross-modal interactions, especially under the evolving dynamics of user intent. Moreover, they frequently neglect modality imbalance issues, leading to suboptimal utilization of multimodal information. To address these challenges, we propose DuAF-MAT, a novel framework for robust multimodal sequential recommendation. Our approach consists of three key components: (1) a Dual-Aware Adaptive Fusion (DuAF) module dynamically calibrates modality contributions by jointly modeling user preferences and temporal information, enabling the extraction of multimodal features aligned with evolving user interests; (2) by integrating Modality Adversarial Training with the Mixture-of-Experts paradigm, MAT-MoE employs an ensemble of expert generators to dynamically reconstruct missing modality representations, effectively mitigating modality imbalance challenges; (3) to address the inherent sparsity of sequential behavior data, we propose a Multi-Supervised Contrastive Learning strategy that integrates cross-modal alignment and virtual sequence augmentation. This approach enhances user interest modeling by leveraging diverse learning signals, resulting in improved model robustness and generalization capability. Extensive experiments on four public datasets demonstrate that DuAF-MAT significantly outperforms state-of-the-art baselines.

Classroom exercises are imperative for reinforcing learning. However, in conventional instruction, students frequently lack timely and personalized feedback. To address this, we present TacpAgent(Teaching Agent for Classroom Practice),a generative LLM-based agent that delivers detailed, individualized guided feedback to promote self-reflection through a prompt-template framework.In the context of this study, Classroom Practice is defined as the set of structured classroom exercises used for formative assessment.TacpAgent takes as input the teacher-prepared exercises and reference answers, together with students’ submitted responses, confidence levels, and brief explanations, then leverages the DeepSeek LLM with designed prompt templates to generate guided feedback and recommend relevant textbook sections for targeted review. We conducted a three-month quasi-experimental study with two high school classes (N=87). The study compared TacpAgent-supported exercises with traditional paper-based exercises. The experimental group showed significantly higher quiz scores (F=18.516, p＜0.001) and improved emotional (p＜0.001) and behavioral engagement (p=0.003). In contrast, the control group demonstrated no significant changes. The results suggest that TacpAgent may enable scalable, personalized formative assessment in classroom settings and provide practical guidance for integrating generative AI into everyday teaching.

ZhongDong Li, Weijie Shi, Yue Cui, Haolun MA, Yuanjun Liu, Jiawei Li, An Liu, Jia Zhu, Jiajie Xu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Yidan Liang, Jia Zhu, Weijie Shi, Hanghui Guo, Yue Cui, Jiawei Shen, Guoqing Ma, Jingjiang Liu, Qingyu Niu, Yilin Wang, Shimin Di, Jiajie Xu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Jingjiang Liu, Jia Zhu, Hanghui Guo, Weijie Shi, Yue Cui, Xiaokang Jin, Yilin Wang, Qingyu Niu, Jiawei Shen, Guoqing Ma, Yidan Liang, Shimin Di, Jiajie Xu. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Cellular Trajectory Map-Matching (CTMM) aims to align cellular location sequences to road networks, which is a necessary preprocessing in location-based services on web platforms like Google Maps, including navigation and route optimization. Current approaches mainly rely on ID-based features and region-specific data to learn correlations between cell towers and roads, limiting their adaptability to unexplored areas. To enable high-accuracy CTMM without additional training in target regions, Zero-shot CTMM requires to extract not only region-adaptive features, but also sequential and location uncertainty to alleviate positioning errors in cellular data. In this paper, we propose a pixel-based trajectory calibration assistant for zero-shot CTMM, which takes advantage of transferable geospatial knowledge to calibrate pixelated trajectory, and then guide the path-finding process at the road network level. To enhance knowledge sharing across similar regions, a Gaussian mixture model is incorporated into VAE, enabling the identification of scenario-adaptive experts through soft clustering. To mitigate high positioning errors, a spatial-temporal awareness module is designed to capture sequential features and location uncertainty, thereby facilitating the inference of approximate user positions. Finally, a constrained path-finding algorithm is employed to reconstruct the road ID sequence, ensuring topological validity within the road network. This process is guided by the calibrated trajectory while optimizing for the shortest feasible path, thus minimizing unnecessary detours. Extensive experiments demonstrate that our model outperforms existing methods in zero-shot CTMM by 16.8\%.

The proliferation of Massive Open Online Courses (MOOCs) has created an urgent need for advanced course recommendation systems (RS). Course recommendations in MOOCs require transparent motivations to justify course selection, as there are often many courses with the same title, but which vary widely in content, duration, learning resources provided, and the academic authority of the instructor. Explainable recommendations are crucial to ensure that recommended courses fit well with learners’ needs and increase the chance of successful course completion, but unfortunately existing RS for MOOCs struggle to provide explainable recommendations. In this article, we present KnowPath , a novel RS for MOOCs, which generates effective and explainable recommendations. KnowPath uses open source Large Language Models (LLMs) to construct knowledge graphs (KGs) capable of accurately capturing complex relationships between MOOC entities (e.g., learners, instructors, educational resources) and employs Reinforcement Learning to align the output of an LLM with learner preferences. Extensive experiments on two public datasets (XueTang and COCO) demonstrate the superior performance and generalizability of KnowPath , underlining its potential to revolutionize the field of personalized online education.

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.

The Dynamic Retrieval Augmented Generation (RAG) paradigm actively decides when and what to retrieve during the text generation process of Large Language Models (LLMs). However, current dynamic RAG methods fall short in both aspects: identifying the optimal moment to activate the retrieval module and crafting the appropriate query once retrieval is triggered. To overcome these limitations, we introduce an approach, namely, RaDIO, Real-Time Hallucination Detection with Contextual Index Optimized query formulation for dynamic RAG. The approach is specifically designed to make decisions on when and what to retrieve based on the LLM’s real-time information needs during the text generation process. We evaluate RaDIO along with existing methods comprehensively over several knowledge-intensive generation datasets. Experimental results show that RaDIO achieves superior performance on all tasks, demonstrating the effectiveness of our work.

Dynamic Retrieval-augmented Generation (RAG) has shown great success in mitigating hallucinations in large language models (LLMs) during generation.However, existing dynamic RAG methods face significant limitations in two key aspects: 1) Lack of an effective mechanism to control retrieval triggers, and 2) Lack of effective scrutiny of retrieval content.To address these limitations, we propose an innovative dynamic RAG method, DioR (Adaptive Cognitive Detection and Contextual Retrieval Optimization), which consists of two main components: adaptive cognitive detection and contextual retrieval optimization, specifically designed to determine when retrieval is needed and what to retrieve for LLMs is useful.Experimental results demonstrate that DioR achieves superior performance on all tasks, demonstrating the effectiveness of our work.

Weijie Shi, Han Zhu, Jiaming Ji, Mengze Li, Jipeng Zhang, Ruiyuan Zhang, Jia Zhu, Jiajie Xu, Sirui Han, Yike Guo. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Weijie Shi, Jipeng Zhang, Yaguang Wu, Jingzhi Fang, Shibo Zhang, Yao Zhao, Hao Chen, Ruiyuan Zhang, Yue Cui, Jia Zhu, Sirui Han, Jiajie Xu, Xiaofang Zhou. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

Knowledge graphs (KGs) have gained popularity in many areas, such as question-answering and recommendation systems, because of their robust knowledge representation capabilities. Real-world KGs are generally incomplete, and therefore significant research efforts have been devoted to finding effective ways to extend the knowledge encapsulated in a KG. Recently, large-language models (LLMs) have been used to enrich the representation of entities and relations in a KG. Such a strategy has its limitations, mainly due to the significant computational resources required by LLMs and the need for a custom query to fully exploit the power of the corresponding LLM. In this paper, we present a novel LLM-based approach to the KG completion task. Specifically, we model an KG triple as a text sequence, so that the entities and relations are used as prompts for the LLM. In this way, we can generate more accurate representations of the entities and relations of the KG. Our method is scalable, capable of running on modest hardware platforms, and replaces custom prompts with automatically generated ones. The effectiveness of our approach is demonstrated through experiments on three real-world datasets, where it achieves state-of-the-art performance in crucial tasks such as triple classification and relation prediction.

Short-video misinformation detection has attracted wide attention in the multi-modal domain, aiming to accurately identify the misinformation in the video format accompanied by the corresponding audio. Despite significant advancements, current models in this field, trained on particular domains (source domains), often exhibit unsatisfactory performance on unseen domains (target domains) due to domain gaps. To effectively realize such domain generalization on the short-video misinformation detection task, we propose deep insights into the characteristics of different domains: (1) The detection on various domains may mainly rely on different modalities (i.e., mainly focusing on videos or audios). To enhance domain generalization, it is crucial to achieve optimal model performance on all modalities simultaneously. (2) For some domains focusing on cross-modal joint fraud, a comprehensive analysis relying on cross-modal fusion is necessary. However, domain biases located in each modality (especially in each frame of videos) will be accumulated in this fusion process, which may seriously damage the final identification of misinformation. To address these issues, we propose a new DOmain generalization model via ConsisTency and invariance learning for shORt-video misinformation detection (named DOCTOR), which contains two characteristic modules: (1) We involve the cross-modal feature interpolation to map multiple modalities into a shared space and the interpolation distillation to synchronize multi-modal learning; (2) We design the diffusion model to add noise to retain core features of multi modal and enhance domain invariant features through cross-modal guided denoising. Extensive experiments demonstrate the effectiveness of our proposed DOCTOR model. Our code is publicly available at https://github.com/ghh1125/DOCTOR.

Diverse decoding of large language models is crucial for applications requiring multiple semantically distinct responses, yet existing methods primarily achieve lexical rather than semantic diversity. This limitation significantly constrains Best-of-N strategies, group-based reinforcement learning, and data synthesis. While temperature sampling and diverse beam search modify token distributions or apply n-gram penalties, they fail to ensure meaningful semantic differentiation. We introduce Semantic-guided Diverse Decoding (SemDiD), operating directly in embedding space that balances quality with diversity through three complementary mechanisms: orthogonal directional guidance, dynamic inter-group repulsion, and position-debiased probability assessment. SemDiD harmonizes these competing objectives using adaptive gain functions and constraint optimization, ensuring both quality thresholds and maximal semantic differentiation. Experiments show SemDiD consistently outperforms existing methods, improving Best-of-N coverage by 1.4-5.2% across diverse tasks and accelerating RLHF training convergence by 15% while increasing accuracy by up to 2.1%.

The integration of artificial intelligence has become a hot topic in the field of education. However, current studies primarily focus on personalization for learners, with the aim of accurately modeling learners’ knowledge level based on their learning history and providing better personalized services, while overlooking the needs of educators. In contrast to the focus on personalization of learners, educators place greater emphasis on accurately assessing collective performance and relative differences within a group, which serves as a qualitative measure of the teaching quality. In this study, we investigate collective knowledge tracing (CeKT), a method designed to estimate the average knowledge level of all students in a course based on a sequence of exercises. To achieve this objective, we propose a graph-based solution capable of estimating the average knowledge level solely from the exercise sequence as well as capturing the intrinsic structure of the exercise sequence (i.e., the sequential order of exercises and the repetition of exercises). Through experimental validation, we affirm that our approach enables a precise estimation of the average knowledge mastery of all students given an exercise sequence and also holds distinct value in three applications within the education domain.

Community detection based on information diffusion is a relatively niche direction. This paper proposes a social network community detection model based on the inner-community information diffusion simulation. The model uses the inner-community information diffusion algorithm adjusted on the naive infectious disease algorithm for node representation, including the dynamic spreading parameter and repeatable infected nodes. Model clusters nodes into communities according to their representation. Experiment results prove that our model outperforms previous related models on synthetic and real datasets. We verify the interpretability of the inner-community information diffusion representation algorithm meanwhile.

Stock prediction is an attractive topic in fintech. However, traditional solutions for stock prediction have two drawbacks: (1) Some focus on the temporal patterns of stocks and model each stock as an independent individual but neglect their relations. Some models consider the relations among stocks, but work in a two-step format (i.e., capturing the temporal patterns first and then considering the relation dependency), which makes them complex and inefficient; (2) They model the stock prediction as a regression (predicting stock price) or classification task (predicting stock trend), which cannot optimize the target of investment, i.e., selecting the best stocks from the exchange market with the highest expected revenue in the future. To fully utilize the relations among stocks and achieve the highest revenue, a relation-temporal graph convolutional network (RT-GCN) is proposed. We first model the relations among stocks and their daily features into a relation-temporal graph. Then, we apply RT-GCN and three relation-aware strategies to realize the relation-temporal feature extraction for each stock. Finally, the features are fed for score calculation in a learning-to-rank way, and the stock with the highest score represents the highest investment revenue in the future. Extensive experiments demonstrate the effectiveness and efficiency of our method.

Epilepsy is one of the most common neurological disorders, and there exists a subset of patients with refractory epilepsy that require surgical removal of the epileptogenic foci (EF) area. Studies have shown that high-frequency oscillations (HFOs) in epileptic electroencephalogram signals can be used as an essential biomarker for locating EF. This paper proposes a new method for rapid localization of EF based on the automatic detection of HFOs by waveform feature templates (WFTs). First, the initial screening of HFOs based on Hilbert transform and subsequent rescreening with short-time energy and short-time Fourier transform is performed, and the two screening results are used as the template data set of HFOs. Then, a coarse-grained and fine-grained screening method for detecting HFOs using autocorrelation coefficients and interrelation coefficients as WFT detectors, respectively. Compared with the Hilbert transform detector and other HFOs detector methods proposed at abroad in recent years, the experimental simulations showed that the automatic detector based on WFT could detect HFOs more rapidly, accurately, and efficiently. Our proposed WFT detector has the advantages of high specificity, high sensitivity, and high accuracy in locating EF and has a high clinical utility.

Epilepsy is a disease caused by abnormal discharges in the central nervous system. Automatic detection and accurate identification of epileptic seizures based on electroencephalography (EEG) are significant in the clinical diagnosis and treatment of epilepsy. In this paper, we first decompose the patient's EEG signal into multiple intrinsic modal functions (IMFs) using empirical modal decomposition, then compute the mean, standard deviation, fluctuation index, and sample entropy of IMF1, and finally classify them using a fusion algorithm of support vector machine and K-nearest neighbor optimized by particle swarm algorithm. The results of validation using the epileptic EEG data set from Bonn University show that the auto-detection and fast recognition method proposed in this paper can achieve a high seizure accuracy recognition rate (≥95%) with only a small number of training samples, which has a good clinical application value.

Monitoring corn crop growth status is of great significance to crop production, breeding, and seed production. The Unmanned Aerial Vehicles’ (UAVs) technology makes it possible to use computer vision technology to identify corn growth stage intelligently. A model customized for corn growth status monitoring based on a dense convolutional neural network (CM-CNN) was proposed, including a two-way dense module and a new activation function ELU. The two-way dense module enlarges the receptive field, while the ELU alleviates gradient disappearance and speeds up learning in deep neural networks. Dense architecture concatenates all the previous layer features to enhance feature reuse. The proposed CM-CNN performs well in classifying corn growth stages. Experimental results show that CM-CNN is a state-of-the-art method, with an accuracy of its relevant data up to 99.3%. Compared with other CNN models, viz. AlexNet, ZFNet, VGG, InceptionV3, Xception and ResNet, fewer parameters are in CM-CNN.

We introduce a unified system for SCHOLAT to import information into Encyclopedia, called Semi-automatic Scholar Encyclopedia Generating System, which consists of four modules: website display page, backstage administration module, version comparison and website security module. Scholar Encyclopedia is implemented as a subsystem of Semi-automatic Scholar Encyclopedia Generating System. Scholar Encyclopedia is a scholar information website that has been independently compiled by the scholars and filled in by personal information. Scholar Encyclopedia Import Model consists of three modules: user evaluation, information import and manual review. At present, Semi-automatic Scholar Encyclopedia Generating System has been used, and the beta version of Scholar Encyclopedia has been deployed in our school network center. The online application shows that Scholar Encyclopedia Import Model has good accuracy and practicability.

Community detection is a well-established problem and nontrivial task in complex network analysis. The goal of community detection is to discover community structures in complex networks. In recent years, many existing works have been proposed to handle this task, particularly nonnegative matrix factorization-based method, e.g. HNMF, BNMF, which is interpretable and can learn latent features of complex data. These methods usually decompose the original matrix into two matrixes, in one matrix, each column corresponds to a representation of community and each column of another matrix indicates the membership between overall pairs of communities and nodes. Then they discover the community by updating the two matrices iteratively and learn the shallow feature of the community. However, these methods either ignore the topological structure characteristics of the community or ignore the microscopic community structure properties. In this paper, we propose a novel model, named Modularized Deep NonNegative Matrix Factorization (MDNMF) for community detection, which preserves both the topology information and the instinct community structure properties of the community. The experimental results show that our proposed models can significantly outperform state-of-the-art approaches on several well-known dataset.

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.

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.

Virtual try-on systems under arbitrary human poses have significant application potential, yet also raise extensive challenges, such as self-occlusions, heavy misalignment among different poses, and complex clothes textures. Existing virtual try-on methods can only transfer clothes given a fixed human pose, and still show unsatisfactory performances, often failing to preserve person identity or texture details, and with limited pose diversity. This paper makes the first attempt towards a multi-pose guided virtual try-on system, which enables clothes to transfer onto a person with diverse poses. Given an input person image, a desired clothes image, and a desired pose, the proposed Multi-pose Guided Virtual Try-On Network (MG-VTON) generates a new person image after fitting the desired clothes into the person and manipulating the pose. MG-VTON is constructed with three stages: 1) a conditional human parsing network is proposed that matches both the desired pose and the desired clothes shape; 2) a deep Warping Generative Adversarial Network (Warp-GAN) that warps the desired clothes appearance into the synthesized human parsing map and alleviates the misalignment problem between the input human pose and the desired one; 3) a refinement render network recovers the texture details of clothes and removes artifacts, based on multi-pose composition masks. Extensive experiments on commonly-used datasets and our newly-collected largest virtual try-on benchmark demonstrate that our MG-VTON significantly outperforms all state-of-the-art methods both qualitatively and quantitatively, showing promising virtual try-on performances.

In this demonstration, we present an end-to-end web-aided data imputation prototype system named WebPut. WebPut consults the Web for imputing the missing values in a local database when the traditional inferring-based imputation method has difficulties in getting the right answers. Specifically, WebPut investigates the interaction between the local inferring-based imputation methods and the web-based retrieving methods and shows that retrieving a small number of selected missing values can greatly improve the imputation recall of the inferring-based methods. Besides, WebPut also incorporates a crowd intervention component that can get advice from humans in case that the web-based imputation methods may have difficulties in making the right decisions. We demonstrate, step by step, how WebPut fills an incomplete table with each of its components.

Manipulating person images under diverse poses, which transfers a person from one pose to another desired pose, is an interesting yet challenging task due to large non-rigid spatial deformation. Most existing works fail to preserve the fine-grained appearance consistency along with the pose changes due to the lack of explicit constraints and spatial modeling, leading to unrealistic results with severe artifacts. In this paper, we propose a novel Part-Preserving Generative Adversarial Network (PP-GAN) to achieve good manipulation quality by explicitly enforcing rich structure constraints over generative modeling. PP-GAN is proposed to decompose the challenging spatial transformation of the whole body into fine-grained part-level transformations, which are then integrated via human joint structure constraint. Given arbitrary poses, PP-GAN integrates human joint structure and region-level part cues as inputs to perform explicit generative modeling. Besides, we introduce a parsing-consistent loss to enforce semantic consistency among images with diverse poses, which guides the image synthesis from a semantic perspective. Extensive qualitative and quantitative evaluations on two benchmarks show that our PP-GAN significantly outperforms the state-of-the-art baselines in generating more realistic and plausible image synthesis results. PP-GAN successfully preserves part-level characteristics even for most challenging pose changes while prior works are easy to fail.

Cross-modal retrieval has gained much attention due to the growing demand for enormous multi-modal data in recent years (i.e., image-text or text-image retrieval). In order to alleviate the problem of ignoring the existence of irrelevant information between images and texts, this paper proposes Deep Pairwise Ranking model with multi-label information for Cross-Modal retrieval (DPRCM). DPRCM directly learns a mapping from images and texts to a compact Euclidean space where distances correspond to the similarity measure of images and texts. The bi-triplet loss function in DPRCM reduces the distance between associated images and texts on the common subspace and increases the margin of independent samples. The classification loss function can better utilize the multi-label information to reduce the semantic gap between image features and text descriptions. Experiments on three widely-used datasets show that DPRCM can achieve competitive performance compared to state-of-the-art methods.

Scene text detection is one of the most challenging tasks in many computer vision applications due to the large variety of scene text appearance and the complexity of scene context. In this paper, we propose an end-to-end trainable framework RefineText for multi-oriented scene text detection, which has a strong ability to detect different-scale texts and split them precisely. High-resolution semantic features are first generated by our designed Feature Refinement Module, which refines features progressively at multiple levels of abstraction. Then text regions are densely produced on the high-level semantic features and followed by Non-Maximum Suppression(NMS) to get final detection results. Experiments on benchmark datasets including ICDAR 2015, ICDAR 2013 and MSRA TD500 demonstrate that our proposed method has competitive performance and strong robustness.

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.

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.

Despite the vast amount of research related to Chinese typo detection, we still lack a publicly available benchmark dataset for evaluation. Furthermore, no precise evaluation schema for Chinese typo detection has been defined. In response to these problems: (1) we release a benchmark dataset to assist research on Chinese typo correction; (2) we present an evaluation schema which was adopted in our NLPTEA 2017 Shared Task on Chinese Spelling Check; and (3) we report new improvements to our Chinese typo detection system ACT.

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.

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.

With the development of modern technology, more and more research papers have been published and shared in various digital databases. However, it is time-consuming for researchers to find suitable scholars who study the same research field with them. To address this issue, we focus on proposing a keyword-based scholar recommendation framework, which can help users to advance their research by recommending scholars that align with the users interests. We first utilize keywords that are extracted from abstract to construct a word-word co-occurrence graph in each query. Based on the graph, we propose an approach to find core nodes to solve cold start problem by treating these core nodes as the users interests. We then combine these core nodes and authors to build a bipartite graph, and adopted the PersonalRank algorithm to rank authors based on the bipartite graph. Finally, we design a recommendation evaluation criterion by comparing our recommendation lists with the results lists of Microsoft Academic search. Experimental results show that our recommendation framework can effectively and efficiently generate scholar recommendation in the situation that lack of the citations times, user behavior and other important indicators.

With the development of the Internet, more and more data can be found on texts information. People produce texts information via writing blogs, product reviews, microblogs, film reviews and so on, which contain sentiments or opinions of the writer. User comments usually can reflect their intuitive feelings for a product. We can dig out relatively large value through sentiment analysis for these comments. Sentiment classification, as one of the most important tasks in sentiment analysis for many real world applications, is our main focus in this article. To improve the accuracy of sentiment classification, we propose a deep neural network fusion framework, which is composed of a multi-window CNN-LSTM model and a multi-window CNN-CNN model with the fusion of the probability of two models to generate final output. Experimental results instruct that our framework is quite feasible.

Multi-modal data lies on heterogeneous feature spaces, which brings a significant challenge to cross-modal retrieval. Some works have been proposed to cope with this problem by learning a common subspace. However, previous methods often learn the common subspace by enhancing the relation between embedded features and relevant class labels but ignore the relation between embedded features and irrelevant class labels. Additionally, most methods assume that irrelevant samples are of equal importance. Considering this, we propose to train an optimal common embedding space via cross-modal learning to rank with adaptive listwise constraint (CMAL2R) based on two-branch neural networks. The listwise loss function in CMAL2R adaptively assigns larger margins to harder irrelevant samples, strengthening the relation between embedded features and irrelevant class labels. Experiments on Wikipedia and Pascal datasets demonstrate the effectiveness for bi-directional image-text retrieval.

Despite remarkable advances in image synthesis research, existing works often fail in manipulating images under the context of large geometric transformations. Synthesizing person images conditioned on arbitrary poses is one of the most representative examples where the generation quality largely relies on the capability of identifying and modeling arbitrary transformations on different body parts. Current generative models are often built on local convolutions and overlook the key challenges (e.g. heavy occlusions, different views or dramatic appearance changes) when distinct geometric changes happen for each part, caused by arbitrary pose manipulations. This paper aims to resolve these challenges induced by geometric variability and spatial displacements via a new Soft-Gated Warping Generative Adversarial Network (Warping-GAN), which is composed of two stages: 1) it first synthesizes a target part segmentation map given a target pose, which depicts the region-level spatial layouts for guiding image synthesis with higher-level structure constraints; 2) the Warping-GAN equipped with a soft-gated warping-block learns feature-level mapping to render textures from the original image into the generated segmentation map. Warping-GAN is capable of controlling different transformation degrees given distinct target poses. Moreover, the proposed warping-block is light-weight and flexible enough to be injected into any networks. Human perceptual studies and quantitative evaluations demonstrate the superiority of our Warping-GAN that significantly outperforms all existing methods on two large datasets.

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.

Cancer diseases are caused by combination of genetic, environmental, and lifestyle factors. Therefore, it is difficult for health organization to treat this disease. This study focuses on identifying the landscape of Cancer research by using bibliometric methods and social network analysis methods based on a number of research articles related to Cancer retrieved from PubMed. To deeply understand the landscape of research on this disease, we adopt productivity analysis which consists of author, university/institution, country and frequent MeSH terms analysis. We specifically perform the concept graph-based network analysis by applying four centrality measures and analyzing co-occurrence of MeSH terms. In the end, we propose a method to predict the Rising Star that may be the active researcher in the field of Cancer disease in the next few years. With this method, we can possibly find more academic cooperation via academic social networks. The encouraging results show that our work is highly feasible.

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.

Summary Online social networks (OSNs) have become more and more popular and have attracted a great many users. Friend recommendation, which is one of the important services in OSN, can help users discover their interested friends and alleviate the problem of information overload. However, most of existing recommendation methods only consider either user link or content information and hence are not effective enough to provide high quality recommendations. In this paper, we propose a topic community‐based method via Nonnegative Matrix Factorization (NMF). This method first applies joint NMF model to mine topic communities existing in OSN by combing link and content information. Then it computes user pairwise similarities and makes friends recommendation based on topic communities. Furthermore, this method can be implemented using the MapReduce distributed computing framework. Extensive experiments show that our proposed method not only has better recommendation performance than state‐of‐the‐art methods but also has good scalability to deal with the problem of friend recommendation in large‐sale OSNs. Moreover, the application case demonstrates that it can significantly improve friend recommendation service in the real world OSN. Copyright © 2016 John Wiley & Sons, Ltd.

One of the fundamental questions is how we can enhance our relationships with the people in our social networks. The combination of frequent engagement, deep interaction, and time spent together are what constitute to build a stronger relationship. With this knowledge in mind, it is not surprising to see many scholars claim that social networking sites may help us to re-connect some lost relationships, but has limited ability to strengthen an existing relationship. Surprisingly, from a survey we conducted, we found that most people who use Facebook or scholat.com (a scholar-oriented social networking site) claim that their relationships with some of their friends are improved, even though they do not have any in-depth or personal conversation with them and they can understand their friends better even without any communication. It contradicts with what we have known. It seems that a relationship can be improved with Facebook or scholat.com alone. In this paper, we try to analyze this finding and explain it using communication model and visualized analysis. We conclude how communication model changes as time goes and how it re-defines the meaning of communication.

With the social network being widely used, people would like to use the friends recommendation provided by a social websites. There are lots of methods to make the recommendation results more accurately and efficiently. By considering the feature selection strategy in the stage of data preprocessing, we propose a novel friend recommendation system using a classification model, which formulates the recommendation problem. We compare the performance of four classifiers, and draw a conclusion that our proposed method can get higher accuracy.

News recommendation recently has attracted wide spread research attention because of the fast propagation of information on the Internet. Due to the large volume of information, a recommendation system which can provide the most important and useful information is required. Most of existing researches focus on providing recommendation based on news contents and predict the category of news only, which is inefficient if the news pool is very large or contains a lot of noisy data. In this study, we propose a novel news recommendation system called UBS, which recommends personalized news based on User Behavior Sequence (UBS) with high efficiency. We formulate the mining problem of user behavior sequence for Internet news reading, which can significantly enhance the performance of recommendation. Experimental validation was conducted using real datasets that obtained from news website. The results show that UBS can provide reasonable news recommendation compared to content-based recommendation as well as collaborative filtering.

Web Harvesting enables the enrichment of incomplete data sets by retrieving required information from the Web. However, the ambiguity of instances may greatly decrease the quality of the harvested data, given that any instance in the local data set may become ambiguous when attempting to identify it on the Web. Although plenty of disambiguation methods have been proposed to deal with the ambiguity problems in various settings, none of them are able to handle the instance ambiguity problem in Web Harvesting. In this paper, we propose to do instance disambiguation in Web Harvesting with a novel disambiguation method inspired by the idea of collaborative identity recognition. In particular, we expect to find some common properties in forms of latent shared attribute values among instances in the list, such that these shared attribute values can differentiate instances within the list against those ambiguous ones on the Web. Our extensive experimental evaluation illustrates the utility of collaborative disambiguation for a popular Web Harvesting application, and shows that it substantially improves the accuracy of the harvested data.

Piecewise Linear Representation (PLR) has been a widely used method for approximating data streams in the form of compact line segments. The buffer-based approach to PLR enables a semi-global approximation which relies on the aggregated processing of batches of streamed data so that to adjust and improve the approximation results. However, one challenge towards applying the buffer-based approach is allocating the necessary memory resources for stream buffering. This challenge is further complicated in a multi-stream environment where multiple data streams are competing for the available memory resources, especially in resource-constrained systems such as sensors and mobile devices.

Entity resolution (ER) is a problem that arises in many areas. In most of cases, it represents a task that multiple entities from different sources require to be identified if they refer to the same or different objects because there are not unique identifiers associated with them. In this paper, we propose a model using web pages identification to identify entities and merge those entities refer to one object together. We use a classical name disambiguation problem as case study and examine our model on a subset of digital library records as the first stage of our work. The favorable results indicated that our proposed approach is highly effective. © 2010 Copyright is held by the author/owner(s).