High-quality multi-hop instruction data is critical for enhancing the reasoning capabilities of large language models (LLMs) in complex long-context scenarios, e.g., long-form reasoning. Nevertheless, there is currently a notable scarcity of such datasets within the community, and existing data synthesis approaches typically fail to provide explicit modeling of intermediate reasoning steps, resulting in unverifiable and potentially erroneous samples. To mitigate above issue, we design the Concept-Graph based Multi-hop Instructions Synthesis (CGMIS) framework, which constructs long-form reasoning paths via concept graph traversal and automatically generates verifiable multi-hop data. The CGMIS framework not only guarantees the accuracy and verifiability of the synthesized data but also enables the construction of high-quality multi-hop instruction datasets from arbitrary corpora. Experiments show that fine-tuning with CGMIS-generated data achieves state-of-the-art performance across 13 long-context reasoning tasks on various models, using only 10% of the data volume required by existing methods.

Previous research on multi-party dialogue generation has predominantly leveraged structural information inherent in dialogues to directly inform the generation process.However, the prevalence of colloquial expressions and incomplete utterances in dialogues often impedes comprehension and weakens the fidelity of dialogue structure representations, which is particularly pronounced in multi-party dialogues.In this work, we propose a novel framework DRCR (Discourse coherence and Responseguided Context Rewriting) to improve multiparty dialogue generation through dialogue context rewriting.Specifically, DRCR employs two complementary feedback signals, discourse coherence and response quality, to construct preference data for both context rewriting and response generation.Moreover, we propose a dynamic self-evolution learning method that allows the rewriter and responder to continuously enhance their capabilities through mutual interaction in an iterative training loop.Comprehensive experiments conducted on four multi-party dialogue datasets substantiate the effectiveness of DRCR.

Previous work on cross-document event coreference resolution (CDECR) primarily focused on enhancing semantic coherence between event mentions, largely overlooking the critical aspect of temporal coherence.To address this issue, we propose CohTP, a novel Temporal Cohorence-driven event coreference framework.CohTP explicitly models and enforces temporal constraints by first constructing a temporal event graph via a fine-tuned natural language inference (NLI) model.The graph is then refined using an Edge-Aware GNN to resolve conflicts and partitioned into ordered time segments, where undirected edges group contemporaneous events.Event coreference resolution is subsequently performed within these temporally coherent segments, where event representations are further augmented with temporally consistent contexts.Experiments on the ECB+, GVC, WEC, and ECB+META datasets show that CohTP outperforms several state-ofthe-art baselines.

Cross-Document Coreference Resolution (CDCR) aims to identify and group together mentions of a specific event or entity that occur across multiple documents.In contrast to the within-document tasks, in which event and entity mentions are linked by rich and coherent contexts, cross-document mentions lack such critical contexts, which presents a significant challenge in establishing connections among them.To address this issue, we introduce a novel task Cross-Document Discourse Coherence Enhancement (CD-DCE) to enhance the discourse coherence between two cross-document event or entity mentions.Specifically, CD-DCE first selects coherent texts and then adds them between two cross-document mentions to form a new coherent document.Subsequently, the coherent text is employed to represent the event or entity mentions and to resolve any coreferent mentions.Experimental results on the three popular datasets demonstrate that our proposed method 1 outperforms several state-of-the-art baselines.

Improving the mathematical reasoning capabilities of Large Language Models (LLMs) is critical for advancing artificial intelligence. However, access to extensive, diverse, and high-quality reasoning datasets remains a significant challenge, particularly for the open-source community. In this paper, we propose ScaleQuest, a novel, scalable, and cost-effective data synthesis method that enables the generation of large-scale mathematical reasoning datasets using lightweight 7B-scale models. ScaleQuest introduces a two-stage question-tuning process comprising Question Fine-Tuning (QFT) and Question Preference Optimization (QPO) to unlock the question generation capabilities of problem-solving models. By generating diverse questions from scratch – without relying on powerful proprietary models or seed data – we produce a dataset of 1 million problem-solution pairs. Our experiments demonstrate that models trained on our data outperform existing open-source datasets in both in-domain and out-of-domain evaluations. Furthermore, our approach shows continued performance improvement as the volume of training data increases, highlighting its potential for ongoing data scaling. The extensive improvements observed in code reasoning tasks demonstrate the generalization capabilities of our proposed method. Our work provides the open-source community with a practical solution to enhance the mathematical reasoning abilities of LLMs.

Dialogue discourse parsing aims to identify and analyze discourse relations between the utterances within dialogues. However, linguistic features in dialogues, such as omission and idiom, frequently introduce ambiguities that obscure the intended discourse relations, posing significant challenges for parsers. To address this issue, we propose a Discourse-aware Clarification Module (DCM) to enhance the performance of the dialogue discourse parser. DCM employs two distinct reasoning processes: clarification type reasoning and discourse goal reasoning. The former analyzes linguistic features, while the latter distinguishes the intended relation from the ambiguous one. Furthermore, we introduce Contribution-aware Preference Optimization (CPO) to mitigate the risk of erroneous clarifications, thereby reducing cascading errors. CPO enables the parser to assess the contributions of the clarifications from DCM and provide feedback to optimize the DCM, enhancing its adaptability and alignment with the parser’s requirements. Extensive experiments on the STAC and Molweni datasets demonstrate that our approach effectively resolves ambiguities and significantly outperforms the state-of-the-art (SOTA) baselines.

Generative Reward Models (GenRMs) leverage synthesized Chains of Thought (CoT) to reduce the need for massive labeled data, but this approach introduces risks of overoptimization due to the inability to guarantee the correctness of the CoTs.Identifying and optimizing unexpected behaviors within these synthesized CoT remains a challenge, as it heavily depends on precise annotations of intermediate behavior, similar to process supervision.In this work, we introduce a criteria-based preference tree for reward modeling, where each path in the tree represents a reasoning trajectory based on synthesized criteria.Crucially, each reasoning trajectory can be independently optimized through RL algorithm.These fine-grained process reward signals are derived from the inferencetime computations and predefined rules, eliminating the need for human supervision.In experiments, SyncPL 1 showed significant improvements over baselines on multiple human preference benchmarks.We further demonstrate that synthesized data can be learned using a long CoT format, analogous to an o1-like model, further enhancing performance while keeping stability and efficiency during training.

Goal-oriented proactive dialogue systems are designed to guide user conversations seamlessly towards specific objectives by planning a goal-oriented path. However, previous research has focused predominantly on optimizing these paths while neglecting the inconsistencies that may arise between generated responses and dialogue contexts, including user profiles, dialogue history, domain knowledge, and subgoals. To address this issue, we introduce a model-agnostic two-stage Consistency Reflection and Correction (CRC) framework. Specifically, in the consistency reflection stage, the model is prompted to reflect on the discrepancies between generated responses and dialogue contexts, identifying inconsistencies and suggesting possible corrections. In the consistency correction stage, the model generates responses that are more consistent with the dialogue context based on these reflection results. We conducted experiments on various model architectures with different parameter sizes, including encoder-decoder models (BART, T5) and decoder-only models (GPT-2, DialoGPT, Phi3, Mistral and LLaMA3), and the experimental results on three datasets demonstrate that our CRC framework significantly improves the consistency between generated responses and dialogue contexts.

Previous work on Incomplete Utterance Rewriting (IUR) has primarily focused on generating rewritten utterances based solely on dialogue context, ignoring the widespread phenomenon of coreference and ellipsis in dialogues. To address this issue, we propose a novel framework called TEO (Two-stage approach on Editing Operation) for IUR, in which the first stage generates editing operations and the second stage rewrites incomplete utterances utilizing the generated editing operations and the dialogue context. Furthermore, an adversarial perturbation strategy is proposed to mitigate cascading errors and exposure bias caused by the inconsistency between training and inference in the second stage. Experimental results on three IUR datasets show that our TEO outperforms the SOTA models significantly.

Previous work on empathetic response generation mainly focused on utilizing the speaker’s emotions to generate responses. However, the performance of identifying fine-grained emotions is limited, introducing cascading errors to empathetic response generation. Moreover, due to the conflict between the information in the dialogue history and the recognized emotions, previous work often generated general and uninformative responses. To address the above issues, we propose a novel framework NEC (Non-Emotion-Centric empathetic dialogue generation) based on contrastive learning and context-sensitive entity and social commonsense, in which the frequent replies and sentences with incorrect emotions are punished through contrastive learning, thereby improving the empathy, diversity and information of the responses. The experimental results demonstrate that our NEC enhances the quality of empathetic generation and generates more diverse responses in comparison with the state-of-the-art baselines.The code will be available at https://github.com/huangfu170/NEC-empchat

Multi-party dialogue discourse parsing is an important and challenging task in natural language processing (NLP). Previous studies struggled to fully understand the deep semantics of dialogues, especially when dealing with complex topic interleaving and ellipsis. To address the above issues, we propose a novel model DDPE (Dialogue Discourse Parsing with Explanations) to integrate external knowledge from Large Language Models (LLMs), which consists of three components, i.e., explanation generation, structural parsing, and contrastive learning. DDPE employs LLMs to generate explanatory and contrastive information about discourse structure, thereby providing additional reasoning cues that enhance the understanding of dialogue semantics. The experimental results on the two public datasets STAC and Molweni show that our DDPE significantly outperforms the state-of-the-art (SOTA) baselines.

Previous work on dialogue topic shift detection has primarily focused on shallow local reasoning, overlooking the importance of considering the global historical structure and local details to elucidate the underlying causes of topic shift. To address the above two issues, we introduce the dual-process theory to this task and design a novel Dual-Module Framework DMF (i.e., intuition and reasoning module) for dialogue topic shift detection to emulate this cognitive process. Specifically, the intuition module employs Large Language Models (LLMs) to extract and store the global topic structure of historical dialogue, while the reasoning module introduces a LLM to generate reasoning samples between the response and the most recent topic of historical dialogue, thereby providing local detail explanations for topic shift. Moreover, we distill the dual-module framework into a small generative model to facilitate more precise reasoning. The experimental results on three public datasets show that our DMF outperforms the state-of-the-art baselines.

Document-level event factuality identification (DEFI) assesses the veracity degree to which an event mentioned in a document has happened, which is crucial for many natural language processing tasks. Previous work assesses event factuality by solely relying on the semantic information within a single document, which fails to identify hard cases where the document itself is hallucinative or counterfactual. There is also a pressing need for more suitable data of this kind. To tackle these issues, we construct Factualusion, a novel corpus with hallucination features that can be used not only for DEFI but can also be applied for hallucination evaluation for large language models. We further propose Trucidator, a graph-based framework that constructs intra-document and cross-document graphs and employs a multi-task learning paradigm to acquire more robust node embeddings, leveraging cross-document inference for more accurate identification. Experiments show that our proposed framework outperformed several baselines, demonstrating the effectiveness of our method.

Spatial relation extraction is to identify the spatial positions and dynamic relations between geographical entities within a given text. Previous research has been hindered by two significant challenges: extracting null-role and multi-mention relations. Null-role relations frequently lack the requisite spatial elements to fully express their semantics, while multi-mention relations may confuse models due to the presence of non-coreferent repeated elements within a sentence. To address the two above issues, we propose a prompt-based method that integrates null-role-specific templates and element markers to enhance the encoding process and subsequently improve the model’s discriminative ability. The experimental results on the SpaceEval dataset demonstrate that our proposed model outperforms the SOTA baselines significantly, particularly in terms of its ability to handle null-role and multi-mention relations.

Rumor detection is to identify and verify the authenticity of information to distinguish between true and false statements. For various types of rumors on social media, there is currently a lack of methods that effectively handle either unimodal or multimodal information simultaneously. Additionally, previous methods also suffered from the lack of images, the inconsistencies between the textual and visual modalities, and those misleading images. To address the above issues, we propose a Universal Rumor Detection Method (URDM), which introduces adversarial training using Diffusion and CLIP to enforce modality consistency and adopts Large Language Models (LLMs) and CoT (Chain of Thought) to generate external knowledge. Specifically, to address the issues of the missing modality and the inconsistencies between text and images, we introduce a random time step fine-tuning method for the Diffusion model to improve CLIP’s ability to judge text-image consistency, and enhance Diffusion’s accuracy in generating images based on text. Concurrently, we incorporated the CoT to serve as an extension of external knowledge to address the issue of misleading images. The experimental results on two popular datasets demonstrate that our proposed URDM outperforms the state-of-the-art baselines.

Most previous work on Conversational QueryRewriting employs an end-to-end rewriting paradigm.However, this approach is hindered by the issue of multiple fuzzy expressions within the query, which complicates the simultaneous identification and rewriting of multiple positions.To address this issue, we propose a novel framework ICR (Iterative Clarification and Rewriting), an iterative rewriting scheme that pivots on clarification questions.Within this framework, the model alternates between generating clarification questions and rewritten queries.The experimental results show that our ICR can continuously improve retrieval performance in the clarification-rewriting iterative process, thereby achieving state-of-the-art performance on two popular datasets.

Large Language Models~(LLMs) have demonstrated superior performance across a wide range of tasks, but they still exhibit undesirable errors due to incorrect knowledge learned from the training data. To avoid this, knowledge editing methods emerged to precisely edit the specific model knowledge via efficiently modifying a very small percentage of parameters. However, those methods can lead to the problem of **Specificity Failure**, where the existing knowledge and capabilities are severely degraded due to editing. Our preliminary indicates that Specificity Failure primarily stems from the model's attention heads assigning excessive attention scores to entities related to the edited knowledge, thereby unduly focusing on specific snippets within the context, which we denote as the **Attention Drift** phenomenon. To mitigate such Attention Drift issue, we introduce a simple yet effective method **S**elective **A**ttention **D**rift **R**estriction(**SADR**), which introduces an additional regularization term during the knowledge editing process to restrict changes in the attention weight distribution, thereby preventing undue focus on the edited entity. Experiments on five frequently-used strong LLMs demonstrate the effectiveness of our method, where SADR can significantly mitigate Specificity Failure in the predominant knowledge editing tasks.

Long-context models (LCMs) have shown great potential in processing long input sequences (even more than 100M tokens) conveniently and effectively. With significant progress, recent research has pointed out that LCMs can accurately locate token-level salient information within the context. Yet, the generation performance of these LCMs is far from satisfactory and might result in misaligned responses, such as hallucinations. To enhance the generation capability of LCMs, existing works have investigated the effects of data size and quality for both pre-training and instruction tuning. Though achieving meaningful improvement, previous methods fall short in either effectiveness or efficiency. In this paper, we introduce LOGO (Long cOntext aliGnment via efficient preference Optimization), a training strategy that first introduces preference optimization for long-context alignment. To overcome the GPU memory-bound issue caused by the long sequence, LOGO employs a reference-free preference optimization strategy and adopts a position synthesis method to construct the training data. By training with only 0.3B data on a single 8 x A800 GPU machine for 16 hours, LOGO allows the Llama-3-8B-Instruct-80K model to achieve comparable performance with GPT-4 in real-world long-context tasks while preserving the model's original capabilities on other tasks, e.g., language modeling and MMLU. Moreover, LOGO can extend the model's context window size while enhancing its generation performance.

The proliferation of fake news videos on social media has heightened the demand for credible verification systems. While existing methods focus on detecting false content, generating human-readable explanations for such predictions remains a critical challenge. Current approaches suffer from spurious correlations caused by two key confounders: 1) video object bias, where co-occurring objects entangle features leading to incorrect semantic associations; and 2) explanation aspect bias, where models over-rely on frequent aspects while neglecting rare ones. To address these issues, we propose CIFE, a causal inference framework that disentangles confounding factors to generate unbiased explanations. First, we formalize the problem through a Structural Causal Model (SCM) to identify confounding factors. We then introduce two novel modules: 1) the Interventional Video-Object Detector (IVOD), which employs backdoor adjustment to decouple object-level visual semantics; and 2) the Interventional Explanation Aspect Module (IEAM), which balances aspect selection during multimodal fusion. Extensive experiments on the FakeVE dataset demonstrate the effectiveness of CIFE, which generates more faithful explanations by mitigating object entanglement and aspect imbalance. Our code is available at https://github.com/Lieberk/CIFE.

Graph neural networks play a pivotal role in various location-based applications, showcasing their compelling ability to capture collaborative signals across user check-in sequences. Recent advancements in next POI recommendation have further leveraged spatio-temporal graphs to uncover the transitional and geographical regularities. However, these methods are usually vulnerable due to the presence of data biases in real-life scenarios, which may mislead the model to disproportionately favoring certain POIs. To this end, this paper proposes a new graph debiasing paradigm for POI recommendation, which disentangles causal and bias knowledge within spatio-temporal graphs, allowing for not only the mitigation of bias issues, but also the utilization of causal information from spatial and temporal perspectives. Specifically, to facilitate graph debiasing at its topological level, an adaptive edge mask generator is first designed to explicitly decompose an entangled graph into causal and bias subgraphs. We encourage the stable relationships between the causal subgraph and the prediction, while the bias subgraph targets at the skewed bias distribution. We further enhance the independence between such two parts by employing a causal-bias disagreement regularization to encourage their distribution in separate semantic spaces. In addition, an inter-view contrastive learning module is also applied to maintain the relation discriminability of transitional and geographical representations. Extensive experiments on three real-world datasets demonstrate the superiority of our proposed model on recommendation performance, as well as its robustness against data bias.

This paper concentrates on Document-level Event Factuality Identification (DEFI) that predicts event factuality values from the viewpoint of the document. At present, the shortcomings of previous studies are multi-fold, including data limitation and scarcity, coarsegrained interpretability without span-level factuality clues, no unified model for different datasets. This paper is devoted to address the above problems by building unified Machine Reading Comprehension (MRC) frameworks comprised of both span-extraction and multiple-choice styles, which exploit Heterogeneous Spectral Attention Networks (HSAN) with spectral networks and hypergraph attention networks as the fine-grained encoders, especially for span-level encoding. Moreover, we integrate Transfer Learning (TL) as cross-domain data augmentation to learn more span-level information from classical MRC datasets by source and target adapters. Experimental performance on ExDLEF corpus, which contains both English and Chinese documents, shows that our span-extraction MRC model is superior to several state-of-the-art baselines, and proves the effectiveness of transfer learning under MRC paradigms.

Multiparty dialog discourse parsing (MDDP) aims to identify the links between pairs of utterances and recognize their discourse relations. Previous research has attempted to address data sparsity in discourse parsing through multitask learning, but these efforts often relied on manually annotated fine-grained information, limiting their practical applicability. In this study, we propose dynamic task-adaptive graph transformer with difficulty-aware task scheduling (DTGT-DTS), an innovative multitask approach that enhances discourse parsing by leveraging neighboring tasks like addressee recognition and speaker identification, without requiring additional annotations. These tasks share common discourse links with discourse parsing but also possess distinct private links. To tackle this, we design a dynamic task-adaptive graph transformer (DTGT) that captures shared links between discourse parsing and its neighboring tasks while distinguishing the private links of neighboring tasks. In addition, we develop a difficulty-aware task scheduling (DTS) strategy that promotes multitask learning by dynamically adjusting training priorities based on the relative difficulty of different tasks. Experimental results on two widely used discourse datasets-Molweni (78 245 links and relations) and STAC (12 691 links and relations)-show that our DTGT-DTS model achieves a 6.07% and 5.31% performance improvement in link identification, respectively, and a 7.27% and 6.02% improvement in relation recognition.

Topic segmentation and outline generation strive to divide a document into coherent topic sections and generate corresponding subheadings, unveiling the discourse topic structure of a document. Compared with sentence-level topic structure, the paragraph-level topic structure can quickly grasp and understand the overall context of the document from a higher level, benefitting many downstream tasks such as summarization, discourse parsing, and information retrieval. However, the lack of large-scale, high-quality Chinese paragraph-level topic structure corpora restrained relative research and applications. To fill this gap, we build the Chinese paragraph-level topic representation, corpus, and benchmark in this paper. Firstly, we propose a hierarchical paragraph-level topic structure representation with three layers to guide the corpus construction. Then, we employ a two-stage man-machine collaborative annotation method to construct the largest Chinese Paragraph-level Topic Structure corpus (CPTS), achieving high quality. We also build several strong baselines, including ChatGPT, to validate the computability of CPTS on two fundamental tasks (topic segmentation and outline generation) and preliminarily verified its usefulness for the downstream task (discourse parsing).

Cross-Document Event Coreference Resolution (CD-ECR) is a task of grouping event mentions across multiple documents that refer to the same real-world events. In contrast to within-document event mentions, which are linked by rich, coherent contexts, cross-document event mentions lack such contexts, making it challenging for the model to establish a connection between two event mentions in different documents. To address this issue, we propose a novel mechanism of enhancing discourse coherence to boost CD-ECR. Specifically, we introduce a new task, ECD-CoE (Event-oriented Cross-Document Coherence Enhancement), which selects coherent sentences that form a coherent text for two cross-document event mentions. We then use this coherent text to represent the event mentions and resolve coreferent events. Experimental results on both the ECB+ and GVC datasets indicate that our proposed method outperforms several state-of-the-art baselines.

Rumors on social media can spread rapidly and widely with the help of the Internet characteristics, causing serious negative impacts on social stability and public life. In order to distinguish rumors from non-rumors, most of the existing methods are based on neural units to encode and observe the content of claims, user comments and rumor propagation patterns. However, these methods only consider the event context information in a single conversation thread, ignoring the public opinion (global contextual information) corresponding to the event in the external news environment. Be aware that users are easily distracted by opinion leaders to false facts and induced to make supportive replies on false claims. In order to address the above-mentioned limitation, we propose a Global Structural-Temporal Graph Network (GSTGN) framework. Specifically, we first construct a multi-modal global opinion graph based on the conversation threads belonging to the same event to capture the external public opinion of the target event. Then to enhance representation learning, we design a Structural-Temporal (ST) unit to encode structural and temporal features of the local conversation graph, and utilize the structural feature of the local graph to guide the learning and encoding of the global opinion graph. Experimental results on two public benchmark datasets prove that our GSTGN method achieves better results than other state-of-the-art models.

Although existing fashionable generation methods on Incomplete Utterance Rewriting (IUR) can generate coherent utterances, they often result in the inclusion of irrelevant and redundant tokens in rewritten utterances due to their inability to focus on critical tokens in dialogue context.Furthermore, the limited size of the training datasets also contributes to the insufficient training of the IUR model.To address the first issue, we propose a multi-task learning framework EO-IUR (Editing Operation-guided Incomplete Utterance Rewriting) that introduces the editing operation labels generated by sequence labeling module to guide generation model to focus on critical tokens.Furthermore, we introduce a token-level heterogeneous graph to represent dialogues.To address the second issue, we propose a two-dimensional utterance augmentation strategy, namely editing operation-based incomplete utterance augmentation and LLM-based historical utterance augmentation.The experimental results on three datasets demonstrate that our EO-IUR outperforms previous state-of-the-art (SOTA) baselines in both open-domain and task-oriented dialogue.

Previous studies on multi-party dialogue generation predominantly concentrated on modeling the reply-to structure of dialogue histories, always overlooking the coherence between generated responses and target utterances.To address this issue, we propose a Reinforcement Learning approach emphasizing both Topic and Rhetorical Coherence (RL-TRC).In particular, the topic-and rhetorical-coherence tasks are designed to enhance the model's perception of coherence with the target utterance.Subsequently, an agent is employed to learn a coherence policy, which guides the generation of responses that are topically and rhetorically aligned with the target utterance.Furthermore, three discourse-aware rewards are developed to assess the coherence between the generated response and the target utterance, with the objective of optimizing the policy.The experimental results and in-depth analyses on two popular datasets demonstrate that our RL-TRC significantly outperforms the state-of-the-art baselines, particularly in generating responses that are more coherent with the target utterances.

Multi-modal Rumor Detection (MRD) has emerged as a crucial research hotpot due to the continuous rise in the spread of multi-modal information on the Internet. Existing studies frequently employ traditional single-classifier models, which cannot accurately classify challenging positive samples. Moreover, the interaction of multiple modalities typically involves an additional fusion module, which results in a trade-off between the granularity of modality interaction and the complexity of the fusion modules. To address these issues, we present a model called Prompt-based Visionaware Classification and Generation (PVCG), where we use a generator module for the MRD. Notably, the encoder independently handles modality fusion more finely by including image as a soft prompt in text embeddings. Our evaluations on Fakeddit and Pheme corpus demonstrate that our PVCG outperforms the state-of-the-art baselines, showcasing its superior performance on the MRD task.

Event Factuality Identification (EFI) aims to assess the factual degree of events in texts, which is crucial and fundamental for many downstream tasks of Natural Language Processing (NLP). Document-level Event Factuality Identification (DEFI), as a branch of EFI tasks that focuses on document text, is an important task in NLP. Currently, research on DEFI is often viewed as a supervised classification task which relies on annotated information. For most datasets of DEFI, the sparsity of the data and the uneven distribution of corpora with different labels limit the research progress of the DEFI task. With the emergence of ChatGPT, it has become feasible for data augmentation by using ChatGPT, which is more accurate and fully functional than all previous tools. This precisely helps to address the problems that most datasets on DEFI have, and we can use ChatGPT to improve and expand the dataset. This paper outlines a systematic framework model, DEFI-G, which leverages Large Language Model (LLM) to process materials reported in the document, greatly expanding the original materials. Moreover, based on GAT, we change the connection method between nodes to improve their performance, resulting in improved task accuracy. The results shows that DEFI-G performs better than all previous baselines.

Event temporal relation extraction is to recognize the temporal relations of two events, which is an important task in natural language processing. Previous work only focused on intra-document information, neglecting the introduction of external knowledge. To address this issue, we propose a Chinese event Temporal relation extraction model on Large language model (LLM) predictive information and Prompt (CTLP). Specifically, we first introduce the large language model ChatGPT to provide inferred information for event pairs to enrich the representation of event information, and then use the prompt method to classify the temporal relations. Moreover, with the enriched semantic information provided by LLM, we perform hierarchical mining of single temporal relations and formulate reasonable hierarchical classification criteria. The experimental results on the Chinese ACE2005-extended dataset show that our model CTLP outperforms the SOTA baselines.

With the rapid development of social media platforms and the increasing scale of the social media data, rumor spreaders are increasingly utilizing multimedia content to attract the attention and trust of news consumers. However, only relying on manual identification will consume a lot of manpower, material and financial resources. Therefore, automatic rumor detectors (i.e., using an efficient method to allow machines to automatically identify rumors) were born. However, most existing methods either only consider text features without utilizing information-rich image or social graph features, or only concatenate text features and image or social graph features, failing to fully explore the correlation between modalities, although they show good performance in rumor detection. At the same time, there are few ways to utilize all three modalities simultaneously. In this paper, we propose a novel Graph Attention Network with Cross-Modal Interaction (GANCI) method for Rumor Detection, which concurrently combines text, image and social graph features. Meanwhile, in GANCI, we designed a Feature Interaction Network to interact text, image and social graph features. Extensive experiments demonstrate the superiority of our model in comparison with the state-of-the-art baselines.

Speculation and Negation Identification focuses on the extraction of speculative and negative cues and scopes. Previous work relied on complete syntactic trees or simply fed sentences into pre-trained language models , which were confronted with poor generalization within and across datasets, and the limitations of training samples. Accordingly, we build a complete pipeline framework that firstly detects cues and then extracts scopes, and propose unified Machine Reading Comprehension paradigms for both cue detection and scope resolution on several datasets. To tackle the insufficiency of training sets and produce more useful samples with appropriate amount of lexical and syntactic knowledge, we apply data augmentation integrating lexical and syntactic features for scope resolution. Experimental results show that our model achieves higher performance than baselines on several publicly accessible corpora.

Event coreference resolution (ECR) aims to group event mentions referring to the same realworld event into clusters.Most previous studies adopt the "encoding first, then scoring" framework, making the coreference judgment rely on event encoding.Furthermore, current methods struggle to leverage human-summarized ECR rules, e.g., coreferential events should have the same event type, to guide the model.To address these two issues, we propose a prompt-based approach, CorefPrompt, to transform ECR into a cloze-style MLM (masked language model) task.This allows for simultaneous event modeling and coreference discrimination within a single template, with a fully shared context.In addition, we introduce two auxiliary prompt tasks, event-type compatibility and argument compatibility, to explicitly demonstrate the reasoning process of ECR, which helps the model make final predictions.Experimental results show that our method CorefPrompt 1 performs well in a state-of-the-art (SOTA) benchmark.

Cross-document event coreference resolution (CD-ECR) is a task of clustering event mentions across multiple documents that refer to the same real-world events.Previous studies usually model the CD-ECR task as a pairwise similarity comparison problem by using different event mention features, and consider the highly similar event mention pairs in the same cluster as coreferent.In general, most of them only consider the local context of event mentions and ignore their implicit global information, thus failing to capture the interactions of long-distance event mentions.To address the above issue, we regard discourse structure as global information to further improve CD-ECR.First, we use a discourse rhetorical structure constructor to construct tree structures to represent documents.Then, we obtain shortest dependency paths from the tree structures to represent interactions between event mention pairs.Finally, we feed the above information to a multi-layer perceptron to capture the similarities of event mention pairs for resolving coreferent events.Experimental results on the ECB+ dataset show that our proposed model outperforms several baselines and achieves the competitive performance with the start-of-theart baselines.

Dialogue discourse parsing aims to reflect the relation-based structure of dialogue by establishing discourse links according to discourse relations.To alleviate data sparsity, previous studies have adopted multitasking approaches to jointly learn dialogue discourse parsing with related tasks (e.g., reading comprehension) that require additional human annotation, thus limiting their generality.In this paper, we propose a multitasking framework that integrates dialogue discourse parsing with its neighboring task addressee recognition.Addressee recognition reveals the reply-to structure that partially overlaps with the relation-based structure, which can be exploited to facilitate relationbased structure learning.To this end, we first proposed a reinforcement learning agent to identify training examples from addressee recognition that are most helpful for dialog discourse parsing.Then, a task-aware structure transformer is designed to capture the shared and private dialogue structure of different tasks, thereby further promoting dialogue discourse parsing.Experimental results on both the Molweni and STAC datasets show that our proposed method can outperform the SOTA baselines.The code will be available at https://github.com/yxfanSuda/RLTST.

With the rapid development of social media, rumor detection on social media has become vitally crucial. Multi-modal fusion and representation play an important role in Multi-modal Rumor Detection (MRD). However, few works learn multi-modal invariant feature and discover the multi-modal class distribution with discrimination loss at the same time. In this paper, we propose a Cross-Modal Adversarial Contrastive (CMAC) fusion strategy, in which adversarial learning is used to align the latent feature distribution of text and image, and contrastive learning is used to align the feature distribution among multi-modal samples of the same category. Adversarial and contrastive learning are combined to obtain multi-modal fusion representations with modality in-variance and clear class distributions. Experimental results on two common benchmark datasets show that our approach achieves better results than other advanced models.

Discourse parsing aims to construct a discourse structure tree to reflect the internal structure of a document. Most existing work only considers parsing documents from the paragraph-level or sentence-level granularity, ignoring the inter-action of different levels of granularity. Therefore, we propose a Multi-Granularity Interaction Method (MGIM) that facilitates discourse parsing through bidirectional information interaction at multi-granularity. We first introduce the structural information at the sentence level to boost paragraph-level parsing, and then use the functional pragmatics information at the paragraph level to guide sentence-level parsing. Moreover, we introduce an auxiliary task, discourse functional pragmatics recognition, to improve sentence-level parsing, which can guide sentence-level discourse tree construction from a macro perspective. Meanwhile, since the research field still lacks data for studying unified Chi-nese discourse parsing, we construct a Unified Chinese Discourse TreeBank UCDTB. Experimental results on both the Chinese UCDTB and the English RST-DT demonstrate the effectiveness of our proposed method.

Extracting event temporal relations is an important task of natural language processing. It is even more challenging to extract the temporal relations of events without using annotated auxiliary information, which is time-consuming and expensive. Therefore, we propose an event temporal relation extraction model ETEMA for Chinese text based on BERT and multi-dimensional attention mechanism, using contextual information interaction and tensor matching methods. Specifically, ETEMA uses BERT to mine the semantic information of event sentences, and uses attention and semantic information to interactively combine event information with its contextual information. Experimental results on ACE2005-extended, a corpus of Chinese temporal relations, show that the proposed model ETEMA achieves optimal performance without using any auxiliary annotated information.

With more and more messages in the form of text and image being spread on the Internet, multi-modal rumor detection has become the focus of recent research. However, most of the existing methods simply concatenate or fuse image features with text features, which can not fully explore the interaction between modalities. Meanwhile, they ignore the convergence inconsistency problem between strong and weak modalities, that is, the dominant rumor text modality may inhibit the optimization of image modality. In this paper, we investigate multi-modal rumor detection from a novel perspective, and propose a Multi-modal Graph Interactive Network with Adaptive Gradient (MGIN-AG) to solve the problem of insufficient information mining within and between modalities, and alleviate the optimization imbalance. Specifically, we first construct fine-grained graph for each rumor text or image to explicitly capture the relation between text tokens or image patches in uni-modal. Then, the cross modal interaction graph between text and image is designed to implicitly mine the text-image interaction, especially focusing on the consistency and mutual enhancement between image patches and text tokens. Furthermore, we extract the embedded text in images as an important supplement to improve the performance of the model. Finally, a strategy of dynamically adjusting the model gradient is introduced to alleviate the under optimization problem of weak modalities in the multi-modal rumor detection task. Extensive experiments demonstrate the superiority of our model in comparison with the state-of-the-art baselines.

Conversational discourse parsing aims to construct an implicit utterance dependency tree to reflect the turn-taking in a multi-party conversation. Existing works are generally divided into two lines: graph-based and transition-based paradigms, which perform well for short-distance and long-distance dependency links, respectively. However, there is no study to consider the advantages of both paradigms to facilitate conversational discourse parsing. As a result, we propose a distance-aware multi-task framework DAMT that incorporates the strengths of transition-based paradigm to facilitate the graph-based paradigm from the encoding and decoding process. To promote multi-task learning on two paradigms, we first introduce an Encoding Interactive Module (EIM) to enhance the flow of semantic information between both two paradigms during the encoding step. And then we apply a Distance-Aware Graph Convolutional Network (DAGCN) in the decoding process, which can incorporate the different-distance dependency links predicted by the transition-based paradigm to facilitate the decoding of the graph-based paradigm. The experimental results on the datasets STAC and Molweni show that our method can significantly improve the performance of the SOTA graph-based paradigm on long-distance dependency links.

Document-level Event Factuality Identification (DEFI) predicts the factuality of a specific event based on a document from which the event can be derived, which is a fundamental and crucial task in Natural Language Processing (NLP). However, most previous studies only considered sentence-level task and did not adopt document-level knowledge. Moreover, they modelled DEFI as a typical text classification task depending on annotated information heavily, and limited to the task-specific corpus only, which resulted in data scarcity. To tackle these issues, we propose a new framework formulating DEFI as Machine Reading Comprehension (MRC) tasks considering both Span-Extraction (Ext) and Multiple-Choice (Mch). Our model does not employ any other explicit annotated information, and utilizes Transfer Learning (TL) to extract knowledge from universal large-scale MRC corpora for cross-domain data augmentation. The empirical results on DLEFM corpus demonstrate that the proposed model outperforms several state-of-the-arts.

Extracting spatial relations from texts is a fundamental task for natural language understanding and previous studies only regard it as a classification task, ignoring those spatial relations with null roles due to their poor information. To address the above issue, we first view spatial relation extraction as a generation task and propose a novel hybrid model HMCGR for this task. HMCGR contains a generation and a classification model, while the former can generate those null-role relations and the latter can extract those non-null-role relations to complement each other. Moreover, a reflexivity evaluation mechanism is applied to further improve the accuracy based on the reflexivity principle of spatial relation. Experimental results on SpaceEval show that HMCGR outperforms the SOTA baselines significantly.

Event coreference resolution (ECR) aims to cluster event mentions that refer to the same real-world events.Deep learning methods have achieved SOTA results on the ECR task.However, due to the encoding length limitation, previous methods either adopt classical pairwise models based on sentence-level context or split each document into multiple chunks and encode them separately.They failed to capture the interactions and contextual cues among those long-distance event mentions.Besides, highlevel information, such as event topics, is rarely considered to enhance representation learning for ECR.To address the above two issues, we first apply a Longformer-based encoder to obtain the document-level embeddings and an encoder with a trigger-mask mechanism to learn sentence-level embeddings based on local context.In addition, we propose an event topic generator to infer the latent topic-level representations.Finally, using the above event embeddings, we employ a multiple tensor matching method to capture their interactions at the document, sentence, and topic levels.Experimental results on the KBP 2017 dataset show that our model 1 outperforms the SOTA baselines.

Document-level Event Factuality Identification (DEFI) predicts the event factuality according to the current document, and mainly depends on event-related tokens and sentences. However, previous studies relied on annotated information and did not filter irrelevant and noisy texts. Therefore, this paper proposes a novel end-to-end model, i.e., Reinforced Multi-Granularity Hierarchical Attention Network (RMHAN), which can learn information at different levels of granularity from tokens and sentences hierarchically. Moreover, with hierarchical reinforcement learning, RMHAN first selects relevant and meaningful tokens, and then selects useful sentences for document-level encoding. Experimental results on DLEF-v2 corpus show that RMHAN model outperforms several state-of-the-art baselines and achieves the best performance.

Sentence-level event factuality identification (SEFI) aims to identify the factuality of an event presented in a sentence. Recent neural network-based approaches have demonstrated the efficacy of the shortest dependency path, but these methods lack semantic information compared with continuous text fragments and may lead to the omission of useful information. In addition, dependency paths are relatively flat. So far, most previous work focused on English datasets, and neglected Chinese tasks. And the existing Chinese SEFI methods ignore the syntactic information. To overcome the above issues, we propose a Chinese event factuality identification model based on dependency trees. We adopt a recursive neural network-based module that fuses event selected predicates, degree words, negative words, and event triggers to capture long-range relations among them. Experimental results on the Chinese event factuality corpus show that our proposed method outperforms other baselines.

Rumors spread through the Internet, especially on Twitter, have harmed social stability and residents’ daily lives. Recently, in addition to utilizing the text features of posts for rumor detection, the structural information of rumor propagation trees has also been valued. Most rumors with salient features can be quickly locked by graph models dominated by cross entropy loss. However, these conventional models may lead to poor generalization, and lack robustness in the face of noise and adversarial rumors, or even the conversational structures that is deliberately perturbed (e.g., adding or deleting some comments). In this paper, we propose a novel Graph Adversarial Contrastive Learning (GACL) method to fight these complex cases, where the contrastive learning is introduced as part of the loss function for explicitly perceiving differences between conversational threads of the same class and different classes. At the same time, an Adversarial Feature Transformation (AFT) module is designed to produce conflicting samples for pressurizing model to mine event-invariant features. These adversarial samples are also used as hard negative samples in contrastive learning to make the model more robust and effective. Experimental results on three public benchmark datasets prove that our GACL method achieves better results than other state-of-the-art models.

Hierarchically constructing micro (i.e., intra-sentence or inter-sentence) discourse structure trees using explicit boundaries (e.g., sentence and paragraph boundaries) has been proved to be an effective strategy. However, it is difficult to apply this strategy to document-level macro (i.e., inter-paragraph) discourse parsing, the more challenging task, due to the lack of explicit boundaries at the higher level. To alleviate this issue, we introduce a topic segmentation mechanism to detect implicit topic boundaries and then help the document-level macro discourse parser to construct better discourse trees hierarchically. In particular, our parser first splits a document into several sections using the topic boundaries that the topic segmentation detects. Then it builds a smaller and more accurate discourse sub-tree in each section and sequentially forms a whole tree for a document. The experimental results on both Chinese MCDTB and English RST-DT show that our proposed method outperforms the state-of-the-art baselines significantly.

As an important research issue in affective computing community, multi-modal emotion recognition has become a hot topic in the last few years. However, almost all existing studies perform multiple binary classification for each emotion with focus on complete time series data. In this paper, we focus on multi-modal emotion recognition in a multi-label scenario. In this scenario, we consider not only the label-to-label dependency, but also the feature-to-label and modality-to-label dependencies. Particularly, we propose a heterogeneous hierarchical message passing network to effectively model above dependencies. Furthermore, we propose a new multi-modal multi-label emotion dataset based on partial time-series content to show predominant generalization of our model. Detailed evaluation demonstrates the effectiveness of our approach.

Implicit discourse relation recognition (IDRR)is a critical task in discourse analysis.Previous studies only regard it as a classification task and lack an in-depth understanding of the semantics of different relations.Therefore, we first view IDRR as a generation task and further propose a method joint modeling of the classification and generation.Specifically, we propose a joint model, CG-T5, to recognize the relation label and generate the target sentence containing the meaning of relations simultaneously.Furthermore, we design three target sentence forms, including the question form, for the generation model to incorporate prior knowledge.To address the issue that large discourse units are hardly embedded into the target sentence, we also propose a target sentence construction mechanism that automatically extracts core sentences from those large discourse units.Experimental results both on Chinese MCDTB and English PDTB datasets show that our model CG-T5 achieves the best performance against several state-of-the-art systems.

Event Argument Extraction aims to extract various arguments of a specific event and assign the roles to them. Currently, most of previous work focused on sentence-level task, ignoring the whole contexts of the event mentions. Obviously, the event mentions in a discourse always are around a specific topic and have interior relations, especially those intra-sentence or adjacent-sentence event mentions. Therefore, this paper proposes an event argument extraction graph model on discourse level. It adds various event mentions, entity mentions and partial semantic dependencies into a graph, which aggregates the discourse contexts by event information, entity information and semantic dependency. The experimental results on the ACE2005 Chinese corpus show that our proposed model outperforms several strong baselines.

Macro discourse relation recognition is an important task of macro discourse analysis. The existing models ignore the micro discourse structure within paragraphs and could not accurately grasp the paragraph semantics. In addition, the traditional pre-trained models only used the representation of the [CLS] token for macro relation classification, lacking more detailed semantic interaction between discourse units. To solve the above issues, we proposed a macro discourse relation recognition model based on Micro Discourse Structure and Self-Interactive Network (MDSSIN) that mines the semantic representation of important parts within a paragraph and enhances semantic interaction between paragraphs. Specially, we first automatically build the micro-structure discourse tree for each paragraph and get the core clause in each paragraph according to nuclearity. Then, we use the self-interactive attention mechanism to capture more detailed semantic interaction between discourse units and between the core clauses of discourse units. Experimental results on Chinese MCDTB show that our method achieves the SOTA performance.

Recognizing discourse relations is crucial for understanding semantic and logical connections between two discourse units in the text. The most of existing work only considers the single-level and single-perspective discourse relationship between discourse units, which will lead to inconsistencies between recognizing classes and sub-classes of the discourse relations and cannot distinguish very similar relations in semantics. Therefore, in this paper, we propose a Multi-perspective and Hierarchical Model (MHM) that can model hierarchical relationships and capture the implicit connections between discourse units from multi-perspective (i.e., rhetorical, co-referential, and temporal) to strengthen the distinction between different discourse relations. Specifically, we first build a hierarchical classification module with contrastive learning to mine the finer semantic in two-level discourse relations and recognize relations consistently. Then, we introduce recognizing co-referential and temporal relations as auxiliary tasks and build a mapping from rhetorical relations to them, modeling multi-perspective relationships between discourse units. The experimental results on MCDTB and CDTB show that our model achieves the best performance.

Discourse structure tree construction is the fundamental task of discourse parsing and most previous work focused on English.Due to the cultural and linguistic differences, existing successful methods on English discourse parsing cannot be transformed into Chinese directly, especially in paragraph level suffering from longer discourse units and fewer explicit connectives.To alleviate the above issues, we propose two reading modes, i.e., the global backward reading and the local reverse reading, to construct Chinese paragraph level discourse trees.The former processes discourse units from the end to the beginning in a document to utilize the left-branching bias of discourse structure in Chinese, while the latter reverses the position of paragraphs in a discourse unit to enhance the differentiation of coherence between adjacent discourse units.The experimental results on Chinese MCDTB demonstrate that our model outperforms all strong baselines.

As an important research issue in the natural language processing community, multi-label emotion detection has been drawing more and more attention in the last few years. However, almost all existing studies focus on one modality (e.g., textual modality). In this paper, we focus on multi-label emotion detection in a multi-modal scenario. In this scenario, we need to consider both the dependence among different labels (label dependence) and the dependence between each predicting label and different modalities (modality dependence). Particularly, we propose a multi-modal sequence-to-set approach to effectively model both kinds of dependence in multi-modal multi-label emotion detection. The detailed evaluation demonstrates the effectiveness of our approach.

Through much exploration in the past decade, emotion analysis in conversations was mainly conducted in textual scenario. Nowadays, with the popularization of speech and video communication, academia and industry have become gradually aware of the need in multimodal scenario. Therefore, emotion detection in conversations becomes increasingly hot not only in natural language processing (NLP) community but also in multimodal analysis community. Although previous studies normally argue that the emotion of current utterance in a conversation is much influenced by the content of historical utterances, their speakers and emotions, they model the influence derived from the history to the current utterance at the same granularity (Intra-modal influence). Intuitively, the clues of emotion detection may not exist in the history of the same modality as current utterance, but in the history of other modalities (Inter-modal influence). Besides, previous studies normally model the information propagation as the conversation flow. Intuitively, bidirectional modeling of information propagation in conversations provides rich clues for emotion detection. Therefore, this paper proposes a bidirectional dynamic dual influence network for real-time emotion detection in conversations, which can simultaneously model both intra- and inter-modal influence with bidirectional information propagation for current utterance and its historical utterances. Detailed experiments demonstrate that our approach much advances the state-of-the-art.

Stance detection aims to assign a stance label (i.e., favor or against) to a post towards a specific target. Recently, there is a growing interest in adopting neural models to detect stance of a document. However, most of these works focus on modeling the sequence of words to learn document representation, though other linguistic information, such as sentiment and arguments, are correlated with the stance of document, and may inspire us to explore the stance. In this article, we propose a hierarchical attention neural model to well study various linguistic information to better represent a document via hierarchical linguistic representations. In addition, we propose a hierarchical network with attention mechanism to weight the importance of various kinds of linguistic information, and learn the mutual attention between document and linguistic information. Detail evaluation on two benchmark datasets demonstrates the effectiveness of proposed hierarchical network with attention mechanism.

In the literature, most of the previous studies on English implicit discourse relation recognition only use sentence-level representations, which cannot provide enough semantic information in Chinese due to its unique paratactic characteristics.In this paper, we propose a topic tensor network to recognize Chinese implicit discourse relations with both sentencelevel and topic-level representations.In particular, besides encoding arguments (discourse units) using a gated convolutional network to obtain sentence-level representations, we train a simplified topic model to infer the latent topic-level representations.Moreover, we feed the two pairs of representations to two factored tensor networks, respectively, to capture both the sentence-level interactions and topiclevel relevance using multi-slice tensors.Experimentation on CDTB, a Chinese discourse corpus, shows that our proposed model significantly outperforms several state-of-the-art baselines in both micro and macro F1-scores.

Longxiang Shen, Bowei Zou, Yu Hong, Guodong Zhou, Qiaoming Zhu, AiTi Aw. 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.

In this paper, we propose a novel approach to multimodal sentiment analysis with focus on both textual and acoustic modalities. Especially, we utilize deep reinforcement learning to explore the clause-level structure in an utterance. On the basis, we perform multimodal interactions at clause-level to model hierarchical interactive representation for multimodal senitment analysis. Detailed evaluation on two benchmark datasets demonstrates the great effectiveness of our approach over several state-of-the-art baselines.

As an emerging research area in natural language processing, multi-modal human language analysis spans language, vision and audio modalities. Understanding multi-modal language requires not only the modeling of independent dynamics within each modality (intra-modal dynamics), but also more importantly interactive dynamics among different modalities (inter-modal dynamics). In this paper, we propose a hierarchical approach to multi-modal language analysis with two levels of attention mechanism, namely interaction-level, which captures the intra-modal and inter-modal dynamics across different modalities with multiple types of attention, and selection-level attention, which selects the effective representations for final prediction by calculating the importance of each vector obtained from interaction-level. Empirical evaluation demonstrates the effectiveness of our proposed approach to multi-modal sentiment classification, sentiment regression and emotion recognition.

Recently, emotion detection in conversations becomes a hot research topic in the Natural Language Processing community. In this paper, we focus on emotion detection in multi-speaker conversations instead of traditional two-speaker conversations in existing studies. Different from non-conversation text, emotion detection in conversation text has one specific challenge in modeling the context-sensitive dependence. Besides, emotion detection in multi-speaker conversations endorses another specific challenge in modeling the speaker-sensitive dependence. To address above two challenges, we propose a conversational graph-based convolutional neural network. On the one hand, our approach represents each utterance and each speaker as a node. On the other hand, the context-sensitive dependence is represented by an undirected edge between two utterances nodes from the same conversation and the speaker-sensitive dependence is represented by an undirected edge between an utterance node and its speaker node. In this way, the entire conversational corpus can be symbolized as a large heterogeneous graph and the emotion detection task can be recast as a classification problem of the utterance nodes in the graph. The experimental results on a multi-modal and multi-speaker conversation corpus demonstrate the great effectiveness of the proposed approach.

Computational modeling of human spoken language is an emerging research area in multimedia analysis spanning across the text and acoustic modalities. Multi-modal sentiment analysis is one of the most fundamental tasks in human spoken language understanding. In this paper, we propose a novel approach to selecting effective sentiment-relevant words for multi-modal sentiment analysis with focus on both the textual and acoustic modalities. Unlike the conventional soft attention mechanism, we employ a deep reinforcement learning mechanism to perform sentiment-relevant word selection and fully remove invalid words of each modality for multi-modal sentiment analysis. Specifically, we first align the raw text and audio at the word level and extract independent handcraft features for each modality to yield the textual and acoustic word sequence. Second, we establish two collaborative agents to deal with the textual and acoustic modalities in spoken language respectively. On this basis, we formulate the sentiment-relevant word selection process in a multi-modal setting as a multi-agent sequential decision problem and solve it with a multi-agent reinforcement learning approach. Detailed evaluations of multi-modal sentiment classification and emotion recognition on three benchmark datasets demonstrate the great effectiveness of our approach over several conventional competitive baselines.

Due to the ability of encoding and mapping semantic information into a highdimensional latent feature space, neural networks have been successfully used for detecting events to a certain extent.However, such a feature space can be easily contaminated by spurious features inherent in event detection.In this paper, we propose a self-regulated learning approach by utilizing a generative adversarial network to generate spurious features.On the basis, we employ a recurrent network to eliminate the fakes.Detailed experiments on the ACE 2005 and TAC-KBP 2015 corpora show that our proposed method is highly effective and adaptable.

Discourse parsing is a challenging task and plays a critical role in discourse analysis. This paper focus on the macro level discourse structure analysis, which has been less studied in the previous researches. We explore a macro discourse structure presentation schema to present the macro level discourse structure, and propose a corresponding corpus, named Macro Chinese Discourse Treebank. On these bases, we concentrate on two tasks of macro discourse structure analysis, including structure identification and nuclearity recognition. In order to reduce the error transmission between the associated tasks, we adopt a joint model of the two tasks, and an Integer Linear Programming approach is proposed to achieve global optimization with various kinds of constraints.

In view of the differences between the annotations of micro and macro discourse rela-tionships, this paper describes the relevant experiments on the construction of the Macro Chinese Discourse Treebank (MCDTB), a higher-level Chinese discourse corpus. Fol-lowing RST (Rhetorical Structure Theory), we annotate the macro discourse information, including discourse structure, nuclearity and relationship, and the additional discourse information, including topic sentences, lead and abstract, to make the macro discourse annotation more objective and accurate. Finally, we annotated 720 articles with a Kappa value greater than 0.6. Preliminary experiments on this corpus verify the computability of MCDTB.

Stance detection aims to assign a stance label (for or against) to a post toward a specific target. Recently, there is a growing interest in using neural models to detect stance of documents. Most of these works model the sequence of words to learn document representation. However, much linguistic information, such as polarity and arguments of the document, is correlated with the stance of the document, and can inspire us to explore the stance. Hence, we present a neural model to fully employ various linguistic information to construct the document representation. In addition, since the influences of different linguistic information are different, we propose a hierarchical attention network to weigh the importance of various linguistic information, and learn the mutual attention between the document and the linguistic information. The experimental results on two datasets demonstrate the effectiveness of the proposed hierarchical attention neural model.

The task of nuclearity recognition in Chinese discourse remains challenging due to the demand for more deep semantic information. In this paper, we propose a novel text matching network (TMN) that encodes the discourse units and the paragraphs by combining Bi-LSTM and CNN to capture both global dependency information and local n-gram information. Moreover, it introduces three components of text matching, the Cosine, Bilinear and Single Layer Network, to incorporate various similarities and interactions among the discourse units. Experimental results on the Chinese Discourse TreeBank show that our proposed TMN model significantly outperforms various strong baselines in both micro-F1 and macro-F1.

Event factuality identification is an important semantic task in NLP. Traditional research heavily relies on annotated texts. This paper proposes a two-step framework, first extracting essential factors related with event factuality from raw texts as the input, and then identifying the factuality of events via a Generative Adversarial Network with Auxiliary Classification (AC-GAN). The use of AC-GAN allows the model to learn more syntactic information and address the imbalance among factuality values. Experimental results on FactBank show that our method significantly outperforms several state-of-the-art baselines, particularly on events with embedded sources, speculative and negative factuality values.

Previous studies on temporal relation extraction focus on mining sentence-level information or enforcing coherence on different temporal relation types among various event mentions in the same sentence or neighboring sentences, largely ignoring those discourse-level temporal relations in nonadjacent sentences. In this paper, we propose a discourse-level global inference model to mine those temporal relations between event mentions in document-level, especially in nonadjacent sentences. Moreover, we provide various kinds of discourse-level constraints, which derived from event semantics, to further improve our global inference model. Evaluation on a Chinese corpus justifies the effectiveness of our discourse-level global inference model over two strong baselines.

Although several semi-supervised learning models have been proposed for English event extraction, there are few successful stories in Chinese due to its special characteristics. In this article, we propose a novel minimally supervised model for Chinese event extraction from multiple views. Besides the traditional pattern similarity view (PSV), a semantic relationship view (SRV) is introduced to capture the relevant event mentions from relevant documents. Moreover, a morphological structure view (MSV) is incorporated to both infer more positive patterns and help filter negative patterns via morphological structure similarity. An evaluation of the ACE 2005 Chinese corpus shows that our minimally supervised model significantly outperforms several strong baselines.

Bowei Zou, Qiaoming Zhu, Guodong Zhou. Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015.

Due to the commonality in natural language, negation focus plays a critical role in deep understanding of context.However, existing studies for negation focus identification major on supervised learning which is timeconsuming and expensive due to manual preparation of annotated corpus.To address this problem, we propose an unsupervised word-topic graph model to represent and measure the focus candidates from both lexical and topic perspectives.Moreover, we propose a document-sensitive biased Pag-eRank algorithm to optimize the ranking scores of focus candidates.Evaluation on the *SEM 2012 shared task corpus shows that our proposed method outperforms the state of the art on negation focus identification.

Active learning (AL) has been proven effective to reduce human annotation efforts in NLP.However, previous studies on AL are limited to applications in a single language.This paper proposes a bilingual active learning paradigm for relation classification, where the unlabeled instances are first jointly chosen in terms of their prediction uncertainty scores in two languages and then manually labeled by an oracle.Instead of using a parallel corpus, labeled and unlabeled instances in one language are translated into ones in the other language and all instances in both languages are then fed into a bilingual active learning engine as pseudo parallel corpora.Experimental results on the ACE RDC 2005 Chinese and English corpora show that bilingual active learning for relation classification significantly outperforms monolingual active learning.

Negative expressions are common in natural language text and play a critical role in information extraction. However, the performances of current systems are far from satisfaction, largely due to its focus on intrasentence information and its failure to consider inter-sentence information. In this paper, we propose a graph model to enrich intrasentence features with inter-sentence features from both lexical and topic perspectives. Evaluation on the *SEM 2012 shared task corpus indicates the usefulness of contextual discourse information in negation focus identification and justifies the effectiveness of our graph model in capturing such global information.

Scope detection is a key task in information extraction.This paper proposes a new approach for tree kernel-based scope detection by using the structured syntactic parse information.In addition, we have explored the way of selecting compatible features for different part-of-speech cues.Experiments on the BioScope corpus show that both constituent and dependency structured syntactic parse features have the advantage in capturing the potential relationships between cues and their scopes.Compared with the state of the art scope detection systems, our system achieves substantial improvement.

Software providers exploit economies of scale by migrating traditional isolated tenancy hosted applications to multi-tenancy supporting ones in the Cloud. How to transform a single-tenancy database to multi-tenancy model plays a key role in the vast majority of migrated SaaS (Software as a Service) applications and it also determines the cost paid by the tenants. In this paper, to meet the tenants' cost-saving expectation versus database performance requirement, a mixed multi-tenancy data model and its architecture generating a configurable multi-tenancy data model group are proposed based on three existing data models firstly. Then, due to the analysis on the challenges arising from such a data migration scenario, this paper describes the design and implementation of our migration approach that enables tenants enjoy customizability of different data models with security and scalability of database services. Finally, a real-world case study demonstrates how our data model and migration approach can be used in the SaaS migration.

Motivated by the critical importance of connectives in recognizing discourse relations, we present an unsupervised cross-argument inference mechanism to implicit discourse relation recognition. The basic idea is to infer the implicit discourse relation of an argument pair from a large number of comparable argument pairs, which are automatically retrieved from the web in an unsupervised way. In this way, the inference proceeds from explicit relations to implicit ones via connective as bridge. This kind of pair-to-pair inference is based on the assumption that two argument pairs with high content similarity (i.e. comparable argument pairs) should have similar discourse relationship. Evaluation on PDTB proves the effectiveness of our inference mechanism in implicit relation recognition to the four level-1 relations. It also shows that our mechanism significantly outperforms other alternatives.

This paper focuses on exploring the features of product reviews that satisfy users, by which to improve the automatic helpfulness voting for the reviews on commercial websites. Compared to the previous work, which single-mindedly adopts the textual features to assess the review helpfulness, we propose that user preferences are more explicit clues to infer the opinions of users on the review helpfulness. By using the user-preference based features, we firstly implement a binary helpfulness based review classification system to divide helpful reviews and useless, and on the basis, we secondly build a Ranking SVM based automatic helpfulness voting system (AHV) which rank reviews based on their helpfulness. Experiments used a large scale dataset containing over 34,266 reviews on 1289 products to test the systems, which achieves promising performances with accuracy of up to 0.72 and [email protected] of 0.25, and at least 9% accuracy improvement compared to the textual-feature based helpfulness assessment.

A complete event description is usually scattered over several sentences and documents, so that how to mine a complete event from several documents or event mentions is an issue currently. This paper proposes an event fusion approach to merge a set of event mentions which distributed over several HTML files into a complete event. Firstly it introduced plain features and structured features into the similarity calculation and applied the hierarchical clustering algorithm to cluster event mentions. Then it proposed an event fusion approach based on a ranking model to merge those argument instances with highest ranking rate in each cluster to form a complete event. The experimental result showed that our approach was effective and could achieve higher accuracy than the baseline.

Sentiment classification is a much-researched field that identifies positive or negative emotions in a large number of texts. Most existing studies focus on document-based approaches and documents are represented as bag-of word. Therefore, this feature representation fails to obtain the relation or associative information between words and it can't distinguish different opinions of a sentiment word with different targets. In this paper, we present a dependency tree-based sentence-level sentiment classification approach. In contrast to a document, a sentence just contains little information and a small set of features which can be used for the sentiment classification. So we not only capture flat features (bag-of-word), but also extract structured features from the dependency tree of a sentence. We propose a method to add more information to the dependency tree and provide an algorithm to prune dependency tree to reduce the noisy, and then introduce a convolution tree kernel-based approach to the sentence-level sentiment classification. The experimental results show that our dependency tree-based approach achieved significant improvement, particularly for implicit sentiment classification.

This paper proposes a novel method to detect text similarity over Chinese research papers using MapReduce paradigm. Our approach differs from the state-of-the-art methods in two aspects. First, we extract the key sentences from Chinese research papers by using some heuristic features and then generate 2-tuple, (document id, key phrase), as the representation of the documents. Second, we design 2-phrase MapReduce algorithm to verify the effectiveness of the generated 2-tuple. For evaluation, we compare the proposed method with other approaches on synthetic corpus. Experimental results review that our method much outperforms the state-of-the-art ones on running time performance while guarantee the Jaccard similarity coefficient.

Application layer protocol identification problem is the premise and foundation of network security monitoring, intrusion detection and other network control system. With the gradual development of network applications, new application layer protocol will produce, therefore bringing the difficulty to identify and update protocols. Although current supervised-based learning method can overcome low accuracy and high time complexity issue of traditional method, yet it is limited by the number of labeled data and also cannot adapt to the rapid updating of the application layer protocols, and it is restricted in further promotion. Accordingly, we propose a semi-supervised learning method to solve the above two issues. Firstly, we adopt Affinity Propagation (AP)clustering algorithm to cluster the mixed data which contains a small labeled data and large unlabeled data. Secondly, we use the labeled data to map the clustering result to specific network application. Evaluation shows that the proposed method is effective in both specific network applications and new protocol identification.

This article proposes a new approach to dynamically determine the tree span for tree kernel-based semantic relation extraction between named entities. The basic idea is to employ constituent dependency information in keeping the necessary nodes and their head children along the path connecting the two entities in the syntactic parse tree, while removing the noisy information from the tree, eventually leading to a dynamic syntactic parse tree. This article also explores various entity features and their possible combinations via a unified syntactic and semantic tree framework, which integrates both structural syntactic parse information and entity-related semantic information. Evaluation on the ACE RDC 2004 English and 2005 Chinese benchmark corpora shows that our dynamic syntactic parse tree much outperforms all previous tree spans, indicating its effectiveness in well representing the structural nature of relation instances while removing redundant information. Moreover, the unified parse and semantic tree significantly outperforms the single syntactic parse tree, largely due to the remarkable contributions from entity-related semantic features such as its type, subtype, mention-level as well as their bi-gram combinations. Finally, the best performance so far in semantic relation extraction is achieved via a composite kernel, which combines this tree kernel with a linear, state-of-the-art, feature-based kernel.

Self-Scheduling is a dynamic and adaptive loop scheduling approach to reduce the total execution time for a task running in the cluster or grid environment. Here, a novel scheme - ACSS (Adaptive Chunk Self- Scheduling Scheme) is proposed to achieve the load balancing and reduce the scheduling overhead. That approach runs the data transfer and data processing tasks in parallel by multiple threads mechanism to reduce the waiting time firstly, and then it adjusts scheduling parameters dynamically based on the node’s performance to achieve better load balancing. The experimental results show that our approach can significantly result less total execution time than other similar ones.

In this paper, a statistic language model is put forward to predict the next inputting word to improve the performance of the input method. So this paper constructs a general language model and a user language model, and then combines them into a new language model which was called as dynamic and self-study language model. Using the general language model in our experiment, the average length of input codes (ALIC) is reduced from 2.557 to 2.479 and the hit rate of first characters (HRFC) is also improved from 78.704% to 96.202%. Using the dynamic and self-study language model in our experiment, when the number of inputted Chinese characters is less then 20 thousand, the HRFC increases rapidly, while the ALIC reduces rapidly. And when the number is greater than 20 thousand, the HRFC and ALIC become steady. Thus it’s clear that dynamic and self-study language model performs well in input method. Otherwise, we provide a modified Church-Gale smoothing method to reduce the size of general language model. This method can reduce the size to 5 percent in order to fit the request of handheld device.

Reasonable analysis and corrective measurement for the testing results of Web applications can effectively judge the effect and efficiency of the testing. Therefore, based on the previous work, we propose a new method for testing results analysis and comparison, which uses the semantic label and XML description technique to realize the information separation between data and display in the Web pages, so as to directly compare the testing results and the expected results. Furthermore, combined with the realities, we determine the metric indexes of Web application testing, so as to provide the criterions and guidelines for the evaluations of the Web applications and their testing processes. And we introduce the feedback control mechanism into the development and evolvement of Web applications, so as to further improve the system quality.

The article puts forward an extended model of role-based access control for Web application system (WAS). This model is named as RBAC4WAS, which comes of RABC model and gives the extended formal description. It is a universal model which is able to apply for the Web application integration area to fit for most of the Web application system in authorization management. The article also expounds the design scheme for Web application integration unified authorization platform (UAP) based on RABC4WAS model. And it describes how to use the UAP to integrate the Web application system to realize a single sign-on portal.