Papers by Deyu Zhou

Multi-Relational Probabilistic Event Representation Learning via Projected Gaussian Embedding (2023.findings-acl)

Challenge:	Existing methods to augment training data with counterfactuals fail to handle multi-hop fact verification due to their incapability to preserve complex logical relationships.
Approach:	They propose to augment training data with counterfactuals that alter causal features of the original data by preserving logical relationships.
Outcome:	The proposed method outperforms the baselines and can generate linguistically diverse counterfactuals without disrupting their logical relationships.

Copied to clipboard

Linhai Zhang, Congzhi Zhang, Deyu Zhou

DINER: Debiasing Aspect-based Sentiment Analysis with Multi-variable Causal Inference (2024.findings-acl)

Challenge:	Existing methods for event representation learning ignore relations and uncertainty of events . Experimental results show that the proposed approach outperforms other state-of-the-art baselines on both existing and newly constructed datasets.
Approach:	They propose a novel approach to learning multi-relational probabilistic event embeddings based on contrastive learning.
Outcome:	The proposed method outperforms existing benchmarks on existing and newly constructed datasets.

Copied to clipboard

Jialong Wu, Linhai Zhang, Deyu Zhou, Guoqiang Xu

Robustness in NLP

Challenge:	Existing approaches to debiase ABSA focus on single-variable causal inference . aspect-based sentiment analysis models are prone to learn spurious correlations from annotation biases .
Approach:	They propose a framework based on multivariable causal inference for debiasing ABSA . they propose to model different types of biases based upon different causal intervention methods .
Outcome:	The proposed framework tackles different types of biases based on different intervention methods.

Explainable Depression Detection in Clinical Interviews with Personalized Retrieval-Augmented Generation (2025.findings-acl)

Copied to clipboard

Linhai Zhang, Ziyang Gao, Deyu Zhou, Yulan He

Cross-Domain Fake News Detection based on Dual-Granularity Adversarial Training (2025.coling-main)

Challenge:	Existing systems rely on black-box neural networks, which lack interpretability, which is crucial in mental health contexts.
Approach:	They propose a Retrieval-augmented generation framework for Explainable depression detection that retrieves evidence from clinical interview transcripts, providing explanations for predictions.
Outcome:	The proposed framework retrieves evidence from clinical interview transcripts, providing explanations for predictions.

Copied to clipboard

Wenjie Wei, Yanyue Zhang, Jinyan Li, Panfei Liu, Deyu Zhou

Neural Topic Modeling with Cycle-Consistent Adversarial Training (2020.emnlp-main)

Challenge:	Existing approaches to detect fake news in unseen domains are limited by domain-specific training.
Approach:	They propose a cross-domain fake news detection method based on adversarial training . they use a document-level and entity-level model to generate domain-independent representations .
Outcome:	The proposed method can detect fake news in unseen domains with the help of pre-trained language models.

Copied to clipboard

Xuemeng Hu, Rui Wang, Deyu Zhou, Yuxuan Xiong

Robustness in NLP

Challenge:	Recent advances on deep generative models have attracted significant interest in neural topic modeling.
Approach:	They propose an adversarial-neural topic model which uses Dirichlet prior to capture the semantic patterns in latent topics.
Outcome:	The proposed models outperform competing models on unsupervised/supervised topic modeling and text classification.

Pre-training and Fine-tuning Neural Topic Model: A Simple yet Effective Approach to Incorporating External Knowledge (2022.acl-long)

Copied to clipboard

Linhai Zhang, Xuemeng Hu, Boyu Wang, Deyu Zhou, Qian-Wen Zhang, Yunbo Cao

Emotion Classification by Jointly Learning to Lexiconize and Classify (2020.coling-main)

Challenge:	Recent studies have shown that using external knowledge such as pre-trained word embeddings or pre-train language models only achieved limited performance improvements but with huge computational overhead.
Approach:	They propose to incorporate external knowledge into neural topic modeling by pre-trained word embeddings (PWEs) or pre-train language models (PLMs) they propose to fine-tune the neural topic model on the target dataset and reduce the huge size of training data.
Outcome:	The proposed approach outperforms current state-of-the-art neural topic models and some topic modeling approaches enhanced with PWEs or PLMs on three datasets and greatly reduces the huge size of training data.

Copied to clipboard

Deyu Zhou, Shuangzhi Wu, Qing Wang, Jun Xie, Zhaopeng Tu, Mu Li

Text Classification

Challenge:	Existing approaches to identify emotions in short text are limited and lack coverage and inaccuracies when applied to informal short text.
Approach:	They propose a novel emotional network to jointly learn sentence emotions and construct emotion lexicons which are dynamically adapted to a given context.
Outcome:	The proposed model outperforms several approaches proposed in previous studies and achieves new state-of-the-art on the benchmark Twitter dataset.

Topic-Driven and Knowledge-Aware Transformer for Dialogue Emotion Detection (2021.acl-long)

Copied to clipboard

Lixing Zhu, Gabriele Pergola, Lin Gui, Deyu Zhou, Yulan He

Topic-Aware Evidence Reasoning and Stance-Aware Aggregation for Fact Verification (2021.acl-long)

Challenge:	Emotion detection in dialogues requires the identification of thematic topics underlying a conversation, commonsense knowledge, and the intricate transition patterns between affective states.
Approach:	They propose a Topic-Driven Knowledge-Aware Transformer model that integrates topic representation and commonsense knowledge from ATOMIC for dialogue emotion detection.
Outcome:	The proposed model outperforms state-of-the-art models on four dialogue datasets . it can detect topics which help distinguish emotion categories, the authors show .

Copied to clipboard

Jiasheng Si, Deyu Zhou, Tongzhe Li, Xingyu Shi, Yulan He

STAR: Constraint LoRA with Dynamic Active Learning for Data-Efficient Fine-Tuning of Large Language Models (2024.findings-acl)

Challenge:	Existing methods for fact verification focus on analyzing semantic interaction between claim and evidence but fail to capture their topical consistency . Existing models focus on the aggregation of multiple pieces of evidence without considering their implicit stances to the claim, thereby introducing spurious information.
Approach:	They propose a topic-aware evidence reasoning and stance-again aggregation model that checks topical consistency between claims and evidence.
Outcome:	The proposed model outperforms state-of-the-art models on two benchmark datasets.

Copied to clipboard

Linhai Zhang, Jialong Wu, Deyu Zhou, Guoqiang Xu

Beyond Text: Incorporating Metadata and Label Structure for Multi-Label Document Classification using Heterogeneous Graphs (2021.emnlp-main)

Challenge:	Existing studies show that supervised training is still necessary for complex reasoning tasks.
Approach:	They propose a method to integrate uncertainty-based active learning and LoRA to effectively integrate the two methods.
Outcome:	The proposed approach outperforms baseline models on three reasoning tasks.

Copied to clipboard

Chenchen Ye, Linhai Zhang, Yulan He, Deyu Zhou, Jie Wu

RGAR: Recurrence Generation-augmented Retrieval for Factual-aware Medical Question Answering (2025.findings-emnlp)

Challenge:	Existing methods for multi-label document classification ignore the heterogeneous graphical structures of metadata and labels.
Approach:	They propose a neural network based approach to multi-label document classification that uses two heterogeneous graphs to model metadata and labels.
Outcome:	The proposed approach outperforms state-of-the-art models on two benchmark datasets.

Copied to clipboard

Sichu Liang, Linhai Zhang, Hongyu Zhu, Wenwen Wang, Yulan He, Deyu Zhou

Open Event Extraction from Online Text using a Generative Adversarial Network (D19-1)

Challenge:	Existing retrieval approaches often overlook patient-specific factual knowledge embedded in EHRs . existing retrieval frameworks often overlook this factual information, limiting its effectiveness in clinical decision-making.
Approach:	They propose a recurrence generation-augmented retrieval framework that synergizes factual and conceptual knowledge from dual sources.
Outcome:	The proposed framework improves on factual-aware medical QA benchmarks.

Copied to clipboard

Rui Wang, Deyu Zhou, Yulan He

Neural Topic Modeling by Incorporating Document Relationship Graph (2020.emnlp-main)

Challenge:	Existing approaches to extract structured representations of open-domain events are limited . a recent study shows that the model outperforms the baseline approaches for extracting events from online texts .
Approach:	They propose an event extraction model based on Generative Adversarial Nets which captures latent events with a generator network and a discriminator to distinguish documents reconstructed from latent and original events.
Outcome:	The proposed model outperforms baseline models on two Twitter and a news article datasets.

Copied to clipboard

Deyu Zhou, Xuemeng Hu, Rui Wang

Code Generation From Flowcharts with Texts: A Benchmark Dataset and An Approach (2022.findings-emnlp)

Challenge:	Graph Topic Models (GNNs) capture relationships between graph nodes via message passing . recent research has focused on topic modeling using latent Dirichlet Allocation .
Approach:	They propose a Graph Topic Model (GTM) that captures relationships between graph nodes via message passing.
Outcome:	The proposed model captures the relationships between nodes via message passing . the results demonstrate that the proposed model is effective in generating documents .

Copied to clipboard

Zejie Liu, Xiaoyu Hu, Deyu Zhou, Lin Li, Xu Zhang, Yanzheng Xiang

Programming Languages in NLP

Challenge:	Currently, researchers focus on generating codes from requirement documents.
Approach:	They propose to generate source code from flowcharts with texts instead of directly translating requirements into codes.
Outcome:	The proposed model improves on the baselines by transforming flowcharts into pseudo-code . the proposed model is based on 320 flowchartes with their corresponding source codes .

Relevant Emotion Ranking from Text Constrained with Emotion Relationships (N18-1)

Copied to clipboard

Deyu Zhou, Yang Yang, Yulan He

Denoising Rationalization for Multi-hop Fact Verification via Multi-granular Explainer (2024.findings-emnlp)

Challenge:	Existing methods to detect emotions from text are lexicon-based and learning-based . experimental results show that the proposed framework is better than state-of-the-art methods .
Approach:	They propose to use a multi-label classification problem to generate a ranked list of relevant emotions.
Outcome:	The proposed framework performs better than state-of-the-art methods and multi-label learning methods on two real-world corpora.

Copied to clipboard

Jiasheng Si, Yingjie Zhu, Wenpeng Lu, Deyu Zhou

AdaRewriter: Unleashing the Power of Prompting-based Conversational Query Reformulation via Test-Time Adaptation (2025.emnlp-main)

Challenge:	Existing rationalization methods for multi-hop fact verification lack nuanced composition in the evidence, which leads to noise rationalization.
Approach:	They propose a method to obtain rationale by completely removing subset of input without compromising verification accuracy.
Outcome:	The proposed method outperforms 12 baselines on three multi-hop fact verification datasets.

Copied to clipboard

Yilong Lai, Jialong Wu, Zhenglin Wang, Deyu Zhou

Dialogue Systems & Conversational Agents

Challenge:	Prompting-based conversational query reformulation has emerged as a powerful approach for conversational search, refining ambiguous user queries into standalone search queries.
Approach:	They propose a framework for query reformulation using an outcome-supervised reward model via test-time adaptation.
Outcome:	Experiments on five conversational search datasets show that AdaRewriter significantly outperforms the existing methods across most settings.

Reduce Redundancy Then Rerank: Enhancing Code Summarization with a Novel Pipeline Framework (2024.lrec-main)

Copied to clipboard

Xiaoyu Hu, Xu Zhang, Zexu Lin, Deyu Zhou

Text Generation

Challenge:	Existing code summarization models lack redundant tokens and are plagued by exposure bias.
Approach:	They propose a pipeline framework to reduce redundancy then rerank that eliminates redundant information in code representation space and a re-ranking model to select more suitable summary candidates.
Outcome:	The proposed framework overrides state-of-the-art approaches on six datasets from the CodeSearchNet benchmark.

MA2P: A Meta-Cognitive Autonomous Intelligent Agents Framework for Complex Persuasion (2026.findings-acl)

Copied to clipboard

Dingyi Zhang, Ziqing Zhuang, Linhai Zhang, Ziyang Gao, Deyu Zhou

Neural Topic Modeling based on Cycle Adversarial Training and Contrastive Learning (2023.findings-acl)

Challenge:	Existing approaches to persuasion generate generic or weakly grounded responses even when such cues are identified.
Approach:	They propose a meta-cognitive autonomous intelligent agent framework for complex persuasion that coordinates perception management, mental-state inference, strategy execution, memory maintenance, and performance evaluation.
Outcome:	The proposed framework achieves a higher persuasion success rate than baselines.

Copied to clipboard

Boyu Wang, Linhai Zhang, Deyu Zhou, Yi Cao, Jiandong Ding

From Script to Stage: Automating Experimental Design for Social Simulations with LLMs (2026.findings-acl)

Challenge:	Neural topic models have been widely used to extract common topics across documents.
Approach:	They propose a framework to apply contrastive learning directly to the decoder . they propose 'self-supervised' contrastive loss to make the generator capture similar topic information .
Outcome:	The proposed framework outperforms baselines on four benchmark datasets.

Copied to clipboard

Yuwei Guo, Zihan Zhao, Xiaowei Liu, Xiangning Yu, Qun Ma, Deyu Zhou, Xiao Xue

Challenge:	Xu et al., 2024): multi-agent simulations based on large language models are a new paradigm for social science research . traditional experimental design relies on interdisciplinary expertise and technical barriers . Xiaoping and Xin eli argue that LLM-driven agents are unreliable for rigorous experimental design due to hallucinations and limited verifiability.
Approach:	They propose a framework for multi-agent experiment design based on script generation . Script Composition, Script Finalization, and Actor Generation are the core phases of the framework .
Outcome:	The proposed framework lowers the barrier for social science experimental design and provides scientifically grounded decision support for policy-making.

TECA: A Two-stage Approach with Controllable Attention Soft Prompt for Few-shot Nested Named Entity Recognition (2024.lrec-main)

Copied to clipboard

Yuanyuan Xu, Linhai Zhang, Deyu Zhou

Low-Resource NLP

Challenge:	Existing methods for few-shot nested named entity recognition (NER) ignore relationship between inner and outer entities, which is crucial for fewshot ner.
Approach:	They propose a span-based method with a controllable attention soft prompt for few-shot nested named entity recognition (TECA) the span part identification provides possible entity mentions without an extra filtering module.
Outcome:	The proposed method outperforms baseline models on four benchmark datasets and outperformed competing models on F1-score by 5.62% on ACE04, 5.11% on ace05, 3.41% on KBP2017 and 0.7% on GENIA on the 10-shot setting.

Sentiment Analysis on Streaming User Reviews via Dual-Channel Dynamic Graph Neural Network (2023.emnlp-main)

Copied to clipboard

Xin Zhang, Linhai Zhang, Deyu Zhou

Multimodality

Challenge:	Existing methods for sentiment analysis on user reviews neglect their time-varying characteristics.
Approach:	They propose a dual-channel framework that models temporal user and product dynamics for sentiment analysis.
Outcome:	The proposed framework is superior to existing methods on five real-world datasets.

Plan Dynamically, Express Rhetorically: A Debate-Driven Rhetorical Framework for Argumentative Writing (2025.emnlp-main)

Copied to clipboard

Xueguan Zhao, Wenpeng Lu, Chaoqun Zheng, Weiyu Zhang, Jiasheng Si, Deyu Zhou

Linguistic Theories

Challenge:	Argumentative essay generation (AEG) is a complex task that requires advanced semantic understanding, logical reasoning, and organized integration of perspectives.
Approach:	They propose a debate-driven rhetorical framework for argumentative writing that integrates Bitzer’s rhetorical situation theory to improve logical depth, argumentative diversity, and rhetorical persuasiveness.
Outcome:	The proposed framework improves logical depth, argumentative diversity, and rhetorical persuasiveness over existing state-of-the-art models.

Implicit Sentiment Analysis with Event-centered Text Representation (2021.emnlp-main)

Copied to clipboard

Deyu Zhou, Jianan Wang, Linhai Zhang, Yulan He

SCOPE: Optimizing Key-Value Cache Compression in Long-context Generation (2025.acl-long)

Challenge:	Existing methods for implicit sentiment analysis simply view noun phrases or entities in text as events or indirectly model events with sophisticated models.
Approach:	They propose an event-centric implicit sentiment analysis that utilizes the sentiment-aware event contained in a sentence to infer sentiment polarity.
Outcome:	The proposed model can detect sentiment in sentences without sentiment words and is compared to existing models on a benchmark dataset.

Copied to clipboard

Jialong Wu, Zhenglin Wang, Linhai Zhang, Yilong Lai, Yulan He, Deyu Zhou

Beyond Static Artifacts: An Evolutionary Framework for Synthetic Claim Generation (2026.acl-long)

Challenge:	Excessive compression during the prefill phase impairs comprehension of reasoning tasks . SCOPE is a framework that performs KV cache optimization during the decoding and prefill phases .
Approach:	They propose a framework that performs optimization during the prefill and decoding phases . they propose enabling a sliding strategy to select essential heavy hitters for the decoding phase .
Outcome:	Experiments show that SCOPE can optimize key-value cache for long-context generation tasks . the framework can preserve essential information while minimizing memory usage and transfer .

Copied to clipboard

Yeqing Teng, Jiasheng Si, Shuxia Lin, Linhai Zhang, Weiyu Zhang, Wenpeng Lu, Deyu Zhou, Xiaoming Wu

Challenge:	Existing claim detection benchmarks treat claims as static textual artifacts . current research ignores sociological etiology of how information naturally emerges and mutates .
Approach:	They propose a socially generative framework for synthetic claim generation . they propose utterance, proposition and context-based simulations to capture truth decay .
Outcome:	The proposed paradigm models claims as socially evolving entities . it allows precise simulation of truth decay and intervened propagation with multi-auditor oversight .

Temporal Knowledge Graph Completion with Approximated Gaussian Process Embedding (2022.coling-1)

Copied to clipboard

Linhai Zhang, Deyu Zhou

A Neural Generative Model for Joint Learning Topics and Topic-Specific Word Embeddings (2020.tacl-1)

Challenge:	Existing TKGC methods are based on deterministic vector embeddings, which are not flexible and expressive enough.
Approach:	They propose a method that maps entities and relations to multivariate Gaussian processes by mapping global trends and local fluctuations in TKGs.
Outcome:	The proposed method can predict global trends and local fluctuations in the TKGs and can be optimized on two real-world benchmark datasets.

Copied to clipboard

Lixing Zhu, Yulan He, Deyu Zhou

WebWalker: Benchmarking LLMs in Web Traversal (2025.acl-long)

Challenge:	Experimental results show that the proposed model outperforms word-level embedding methods in word similarity evaluation and word sense disambiguation.
Approach:	They propose a generative model that explores local and global context for joint learning topics and topic-specific word embeddings.
Outcome:	The proposed model outperforms word-level embedding methods in word similarity evaluation and word sense disambiguation.

Copied to clipboard

Jialong Wu, Wenbiao Yin, Yong Jiang, Zhenglin Wang, Zekun Xi, Runnan Fang, Linhai Zhang, Yulan He, Deyu Zhou, Pengjun Xie, Fei Huang

Information Retrieval

Challenge:	Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of natural language processing tasks.
Approach:	They propose a benchmark to assess the ability of LLMs to perform web traversal by using an explore-critic paradigm.
Outcome:	The proposed framework mimics human-like web navigation through an explore-critic paradigm and demonstrates the effectiveness of RAG combined with WebWalker in real-world scenarios.

PROPER: A Progressive Learning Framework for Personalized Large Language Models with Group-Level Adaptation (2025.acl-long)

Copied to clipboard

Linhai Zhang, Jialong Wu, Deyu Zhou, Yulan He

Language Models

Challenge:	Personalized large language models (LLMs) aim to tailor outputs to user preferences . however, user data is typically sparse, making it challenging to adapt LLMs to specific user patterns.
Approach:	They propose a progressive learning framework that groups users based on preferences and adapts LLMs in stages.
Outcome:	The proposed approach outperforms SOTA models across multiple tasks.

Neural Storyline Extraction Model for Storyline Generation from News Articles (N18-1)

Copied to clipboard

Deyu Zhou, Linsen Guo, Yulan He

Focusing, Bridging and Prompting for Few-shot Nested Named Entity Recognition (2023.findings-acl)

Challenge:	Existing approaches to storyline generation are domain dependent and cannot deal with unseen event types.
Approach:	They propose a neural network-based approach to extract structured representations and evolution patterns of storylines without using annotated data.
Outcome:	The proposed model outperforms state-of-the-art approaches on accuracy and efficiency on three news corpora and it is based on supervised models.

Copied to clipboard

Yuanyuan Xu, Zeng Yang, Linhai Zhang, Deyu Zhou, Tiandeng Wu, Rong Zhou

Low-Resource NLP

Challenge:	Existing work on few-shot named entity recognition (NER) addresses flat entities instead of nested entities.
Approach:	They propose a method based on focusing, bridging and prompting for few-shot nested NER without using source domain data.
Outcome:	The proposed method outperforms baseline models on four benchmark datasets and outperformed several competing models on F1-score by 9.33% on ACE2004, 6.17% on ace2005, 9.40% on GENIA and 5.12% on KBP2017.

SynGraph: A Dynamic Graph-LLM Synthesis Framework for Sparse Streaming User Sentiment Modeling (2025.findings-acl)

Copied to clipboard

Xin Zhang, Qiyu Wei, Yingjie Zhu, Linhai Zhang, Deyu Zhou, Sophia Ananiadou

Multimodality

Challenge:	Traditional sentiment analysis methods focus on static reviews, failing to capture temporal relationship between user sentiment rating and textual content.
Approach:	They propose a dynamic graph-based framework that addresses data sparsity in streaming reviews.
Outcome:	The proposed framework reduces data sparsity by categorizing users into mid-tail, long-tail and extreme scenarios and incorporating LLM enhancements within a dynamic graph-based structure.

Interpretable Relevant Emotion Ranking with Event-Driven Attention (D19-1)

Copied to clipboard

Yang Yang, Deyu Zhou, Yulan He, Meng Zhang

Explainability & Interpretability in NLP

Challenge:	Existing studies ignore the latent event information in documents . Existing methods for detecting emotions are limited to a few words .
Approach:	They propose to integrate event information into a deep learning architecture to extract relevant emotion ranking models using corpus-level event embeddings and document-level events.
Outcome:	The proposed model performs better than state-of-the-art emotion detection and multi-label approaches on three real-world corpora and interpretable results shed light on the events which trigger certain emotions.

An Interpretable Neural Network with Topical Information for Relevant Emotion Ranking (D18-1)

Copied to clipboard

Yang Yang, Deyu Zhou, Yulan He

Explainability & Interpretability in NLP

Challenge:	Experimental results show that the extracted emotion-associated topic words represent emotion-evoking events.
Approach:	They propose an interpretable neural network approach for relevant emotion ranking . they initialize the hidden layer to approximate the behavior of topic models .
Outcome:	The proposed approach performs better than state-of-the-art methods on real-world corpora.

Disentangling Text Representation With Counter-Template For Unsupervised Opinion Summarization (2023.findings-acl)

Copied to clipboard

Yanyue Zhang, Deyu Zhou

Complicate Then Simplify: A Novel Way to Explore Pre-trained Models for Text Classification (2022.coling-1)

Challenge:	Existing approaches for unsupervised opinion summarization are based on reconstruction model, but selection is too coarse as not all information in each input is equally essential for the summary.
Approach:	They propose a framework for unsupervised opinion summarization based on text representation disentanglement with counter-template.
Outcome:	The proposed framework outperforms the state-of-the-art models on quality and stability on two benchmark datasets.

Copied to clipboard

Xu Zhang, Zejie Liu, Yanzheng Xiang, Deyu Zhou

Information Retrieval

Challenge:	Existing frameworks for text classification employing pre-trained models are constrained by the difficulty of the task.
Approach:	They propose a framework which implements a two-stage training strategy to fully exploit the knowledge in pre-trained models.
Outcome:	The proposed framework outperforms state-of-the-art classification models on six text classification corpora.

SEE-Few: Seed, Expand and Entail for Few-shot Named Entity Recognition (2022.coling-1)

Copied to clipboard

Zeng Yang, Linhai Zhang, Deyu Zhou

Low-Resource NLP

Challenge:	Existing few-shot named entity recognition methods focus on leveraging existing datasets in the rich-resource domains which might fail in training-from-scratch setting.
Approach:	They propose a multi-task learning framework for Few-shot named entity recognition without using source domain data.
Outcome:	The proposed framework outperforms state-of-the-art few-shot named entity recognition methods on a training-from-scratch dataset.

When KV Cache Reuse Fails in Multi-Agent Systems: Cross-Candidate Interaction is Crucial for LLM Judges (2026.acl-long)

Copied to clipboard

Sichu Liang, Zhenglin Wang, null Chujiajia, Pengfei Xia, Hui Zang, Deyu Zhou

Challenge:	Multi-agent LLMs generate multiple candidate responses that are aggregated by an LLM judge.
Approach:	They propose to advocate KV cache reuse across partially shared contexts and report substantial speedups for generation agents.
Outcome:	The proposed reuse strategies weaken cross-candidate attention, especially for later candidate blocks, and highlight judge-centric inference as a distinct regime that requires dedicated, risk-aware system design.

Rehearse With User: Personalized Opinion Summarization via Role-Playing based on Large Language Models (2025.findings-acl)

Copied to clipboard

Yanyue Zhang, Yulan He, Deyu Zhou

Summarization

Challenge:	Recent studies show that large language models can achieve stateof-the-art performance on standard summarization benchmarks without the need for large-scale training data.
Approach:	They propose a personalized opinion summarization framework via LLM-based role-playing to better understand the user's personalized needs.
Outcome:	The proposed framework can improve the level of personalization in large model-generated summaries by taking into account user characteristics and interests while summarizing multiple product reviews.

A Novel Framework Based on Medical Concept Driven Attention for Explainable Medical Code Prediction via External Knowledge (2022.findings-acl)

Copied to clipboard

Tao Wang, Linhai Zhang, Chenchen Ye, Junxi Liu, Deyu Zhou

Explainability & Interpretability in NLP

Challenge:	Existing methods to predict medical codes from clinical notes lack interpretability due to lengthy and noisy clinical notes.
Approach:	They propose a framework based on medical concept driven attention to integrate external knowledge for explainable medical code prediction from clinical notes.
Outcome:	The proposed framework outperforms state-of-the-art methods on a benchmark dataset showing that it is more accurate than existing methods.

Neural Temporal Opinion Modelling for Opinion Prediction on Twitter (2020.acl-main)

Copied to clipboard

Lixing Zhu, Yulan He, Deyu Zhou

Large Language Models Have Intrinsic Meta-Cognition, but Need a Good Lens (2025.emnlp-main)

Challenge:	Existing studies have used a manual segmentation of a tweet sequence into equallyspaced intervals based on either tweet counts or time duration.
Approach:	They propose to model users’ tweet posting behaviour as a temporal point process to jointly predict the posting time and the stance label of the next tweet given a user’s historical tweet sequence and tweets posted by their neighbours.
Outcome:	The proposed model predicts the posting time and the stance labels of future tweets more accurately compared to baselines.

Copied to clipboard

Ziyang Ma, Qingyue Yuan, Zhenglin Wang, Deyu Zhou

Language Models

Challenge:	Existing studies have focused on the cognitive error detection capabilities of Large Language Models (LLMs), but few studies have examined the meta-cognitive abilities of LLMs.
Approach:	They propose an automated meta-cognition evaluation framework for evaluation of LLMs and a Markovian Intrinsic Reward Adjustment strategy to boost current lenses.
Outcome:	The proposed framework can be used to evaluate the meta-cognition abilities of LLMs and improve them.

A Multi-label Multi-hop Relation Detection Model based on Relation-aware Sequence Generation (2021.findings-emnlp)

Copied to clipboard

Linhai Zhang, Deyu Zhou, Chao Lin, Yulan He

Knowledge Representation

Challenge:	Existing methods treat multi-label learning problem as a single label . Existing approaches focus on measuring semantic similarity of questions and candidate relations .
Approach:	They propose to solve multi-hop relation detection problem by generating sequences of hops and labels.
Outcome:	The proposed method is effective in KBQA, despite the unknown number of labels and hops.

SEED: Accelerating Reasoning Tree Construction via Scheduled Speculative Decoding (2025.coling-main)

Copied to clipboard

Zhenglin Wang, Jialong Wu, Yilong Lai, Congzhi Zhang, Deyu Zhou

Let the Comments Speak: A Multi-Agent Framework based on Large Language Model for Comment-Guided Code Refactoring (2026.findings-acl)

Challenge:	Large Language Models (LLMs) have remarkable emergent abilities across various tasks, yet their performance on complex reasoning and planning tasks remains suboptimal.
Approach:	They propose a tree-search-based reasoning framework that encourages the exploration of intermediate steps and a round-scheduled strategy to manage draft model dispatching.
Outcome:	The proposed framework improves runtime speed and GPU memory management concurrently and handles multiple iterations for thought generation and state evaluation.

Copied to clipboard

Zixuan Wang, Wutong Yu, Deyu Zhou

A Divide-And-Conquer Approach for Multi-label Multi-hop Relation Detection in Knowledge Base Question Answering (2021.findings-emnlp)

Challenge:	Current Large Language Models focus on syntax and ignore the vital semantic signals in code comments.
Approach:	They propose a Multi-Agent framework for COmment-guided code Refactoring that populates original code with precise comments to provide necessary semantic guidance for subsequent refactoring.
Outcome:	The proposed framework significantly improves code quality and achieves higher developer acceptance compared to baselines.

Copied to clipboard

Deyu Zhou, Yanzheng Xiang, Linhai Zhang, Chenchen Ye, Qian-Wen Zhang, Yunbo Cao

Beyond Meta-Reasoning: Metacognitive Consolidation for Self-Improving LLM Reasoning (2026.acl-long)

Challenge:	Existing methods for relation detection only detect one path to obtain the answer without considering other correct paths.
Approach:	They propose a divide-and-conquer approach for multi-label multi-hop relation detection . they propose 'path sampling mechanism' to generate diverse relation paths .
Outcome:	The proposed approach outperforms other competitive approaches on the FreebaseQA benchmark dataset.

Copied to clipboard

Ziqing Zhuang, Linhai Zhang, Jiasheng Si, Deyu Zhou, Yulan He

LASS: A Novel and Economical Data Augmentation Framework Based on Language Models for Debiasing Opinion Summarization (2025.coling-main)

Challenge:	Existing approaches for improving LLM reasoning remain episodic and lack reusable meta-reasoning skills.
Approach:	They propose a framework that consolidates metacognitive experience from past reasoning episodes into reusable knowledge that improves future meta-reasoning.
Outcome:	The proposed framework consolidates metacognitive experience from past reasoning episodes into reusable knowledge that improves future meta-reasoning.

Copied to clipboard

Yanyue Zhang, Pengfei Li, Yilong Lai, Yulan He, Deyu Zhou

AdaCQR: Enhancing Query Reformulation for Conversational Search via Sparse and Dense Retrieval Alignment (2025.coling-main)

Challenge:	Existing methods to generate negative summaries are expensive and lack the capacity to generate large data sets.
Approach:	They propose a data augmentation framework based on LArge and Small language models for debiaSing opinion summarization that generates a small number of synthesized negative reviews by rewriting the positive text via a large language model.
Outcome:	The proposed framework can generate large numbers of negative reviews by rewriting the positive text using a large language model and training a disentangle reconstruction model based on the generated data.

Copied to clipboard

Yilong Lai, Jialong Wu, Congzhi Zhang, Haowen Sun, Deyu Zhou

Opinions Are Not Always Positive: Debiasing Opinion Summarization with Model-Specific and Model-Agnostic Methods (2024.lrec-main)

Challenge:	Existing methods to address conversational search challenges are limited by one specific retrieval system.
Approach:	They propose a framework to enhance generalizability of information-seeking queries by aligning reformulation models with term-based and semantic retrieval systems.
Outcome:	The proposed framework outperforms existing methods in a more efficient framework.

Copied to clipboard

Yanyue Zhang, Yilong Lai, Zhenglin Wang, Pengfei Li, Deyu Zhou, Yulan He

G3R: A Graph-Guided Generate-and-Rerank Framework for Complex and Cross-domain Text-to-SQL Generation (2023.findings-acl)

Challenge:	Existing opinion summarization frameworks are reluctant to generate negative summaries given input of negative opinions.
Approach:	They propose to disentangle input into sentiment-relevant and sentiment-irrelevant components through adversarial loss.
Outcome:	The proposed approaches reduce sentiment bias in the existing opinion summarization dataset . the proposed approaches generate better summaries with a more balanced emotional polarity distribution .

Copied to clipboard

Yanzheng Xiang, Qian-Wen Zhang, Xu Zhang, Zejie Liu, Yunbo Cao, Deyu Zhou

Multimodality

Challenge:	Existing approaches to complex and cross-domain Text-to-SQL generation lack domain knowledge . domain knowledge is not incorporated to enhance their ability to generalise to unseen databases.
Approach:	They propose a framework called G3R for complex and cross-domain Text-to-SQL generation . they propose re-ranking SQL queries based on domain knowledge and a graph-guided SQL generator .
Outcome:	The proposed framework achieves state-of-the-art results on the Spider and Spider-DK benchmarks.

Neural Topic Modeling with Bidirectional Adversarial Training (2020.acl-main)

Copied to clipboard

Rui Wang, Xuemeng Hu, Deyu Zhou, Yulan He, Yuxuan Xiong, Chenchen Ye, Haiyang Xu

Benchmarking Temporal Reasoning and Alignment Across Chinese Dynasties (2026.eacl-short)

Challenge:	Recent studies have shown that neural topic models for automatic topic extraction avoid complicated mathematical derivations for model inference.
Approach:	They propose a bidirectional adversarial topic model which uses a generator and an encoder to infer topic distribution.
Outcome:	The proposed model outperforms baselines and competitive models in three benchmark corpora.

Copied to clipboard

Zhenglin Wang, Jialong Wu, Pengfei Li, Yong Jiang, Deyu Zhou