Papers by Meishan Zhang

When Words Smile: Generating Diverse Emotional Facial Expressions from Text (2025.emnlp-main)

Challenge:	Pretrained language models (PLMs) have been successful in addressing word boundaries in Chinese sequence labeling tasks, but they rarely consider boundary information explicitly.
Approach:	They propose a method to integrate unsupervised boundary information into Chinese BERT's pre-training objectives and a supervised boundary-aware PLM.
Outcome:	The proposed model outperforms the vanilla version on Chinese sequence labeling tasks and in broader Chinese natural language understanding tasks.

Copied to clipboard

Haidong Xu, Meishan Zhang, Hao Ju, Zhedong Zheng, Erik Cambria, Min Zhang, Hao Fei

Challenge:	Existing systems that generate only coarse facial expressions ignore the rich and dynamic nature of face-to-face communication.
Approach:	They propose an end-to-end text-to expression model that explicitly focuses on emotional dynamics.
Outcome:	The proposed model outperforms baselines on 15,000 text–3D expression pairs on a large-scale dataset.

Reflection on Knowledge Graph for Large Language Models Reasoning (2025.findings-acl)

Copied to clipboard

Yigeng Zhou, Wu Li, Yifan Lu, Jing Li, Fangming Liu, Meishan Zhang, Yequan Wang, Daojing He, Honghai Liu, Min Zhang

mGTE: Generalized Long-Context Text Representation and Reranking Models for Multilingual Text Retrieval (2024.emnlp-industry)

Challenge:	Existing methods for supplementing Large Language Models (LLMs) with knowledge graphs often introduce noise in the retrieval and reasoning pipeline, hindering their ability to integrate external knowledge for complex multi-hop question answering.
Approach:	They propose a framework to enhance LLMs' reasoning capabilities through reflective engagement with knowledge graphs by Query Decoupling, LLM-Driven Knowledge Graph Exploration, and Inference with Knowledge Reconstruction.
Outcome:	The proposed framework integrates external knowledge into LLMs and trains them to leverage this knowledge for answering questions.

Copied to clipboard

Xin Zhang, Yanzhao Zhang, Dingkun Long, Wen Xie, Ziqi Dai, Jialong Tang, Huan Lin, Baosong Yang, Pengjun Xie, Fei Huang, Meishan Zhang, Wenjie Li, Min Zhang

Extending Phrase Grounding with Pronouns in Visual Dialogues (2022.emnlp-main)

Challenge:	Existing models for text retrieval are based on a multi-stage process that involves retrieving documents from a large corpus.
Approach:	They propose to build a multilingual text representation model and a cross-encoder reranker from scratch for text retrieval.
Outcome:	The proposed models outperform the state-of-the-art models on long-context retrieval benchmarks.

Copied to clipboard

Panzhong Lu, Xin Zhang, Meishan Zhang, Min Zhang

Visual Spatial Description: Controlled Spatial-Oriented Image-to-Text Generation (2022.emnlp-main)

Challenge:	Conventional phrase grounding aims to localize noun phrases mentioned in a caption to their corresponding image regions.
Approach:	They extend the task by considering pronouns to include noun phrases and pronounos . they construct a dataset of phrase grounding with noun and pronom phrases to image regions .
Outcome:	Experiments show that pronouns are easier to ground than noun phrases . a baseline model with coreference information can significantly boost the grounding performance .

Copied to clipboard

Yu Zhao, Jianguo Wei, ZhiChao Lin, Yueheng Sun, Meishan Zhang, Min Zhang

Beyond Chunking: Discourse-Aware Hierarchical Retrieval for Long Document Question Answering (2026.acl-long)

Challenge:	Image-to-text tasks such as captioning and controllable image descriptions have received extensive attention for decades.
Approach:	They propose a new perspective for image-to-text to generate spatial descriptions by combining two objects in an image.
Outcome:	The proposed model is awe-inspiring and human-like, and the proposed end-to-end architecture is the better choice for their integration.

Copied to clipboard

Huiyao Chen, Yi Yang, Yinghui Li, Meishan Zhang, Baotian Hu, Min Zhang

Natural Language Interfaces

Challenge:	Existing long document question answering systems process texts as flat sequences or use heuristic chunking, which overlooks the discourse structures that guide human comprehension.
Approach:	They propose a discourse-aware hierarchical framework that leverages rhetorical structure theory for long document question answering.
Outcome:	The proposed framework exhibits strong robustness across diverse document types and linguistic settings.

Semantic Role Labeling from Chinese Speech via End-to-End Learning (2024.findings-acl)

Copied to clipboard

Huiyao Chen, Xinxin Li, Meishan Zhang, Min Zhang

Scene Graph as Pivoting: Inference-time Image-free Unsupervised Multimodal Machine Translation with Visual Scene Hallucination (2023.acl-long)

Challenge:	Semantic role labeling (SRL) has traditionally focused on text input.
Approach:	They propose an end-to-end approach for SRL from speech integrating ASR and SRL in a joint-learning framework, focusing on the Chinese language.
Outcome:	The proposed model improves on the Chinese Proposition Bank 1.0 dataset and the existing model with improved performance.

Copied to clipboard

Hao Fei, Qian Liu, Meishan Zhang, Min Zhang, Tat-Seng Chua

Domain-Specific NER via Retrieving Correlated Samples (2022.coling-1)

Challenge:	Current neural machine translation (NMT) relies on parallel sentences, which obstructs the development of NMT for minor languages.
Approach:	They propose an unsupervised multimodal machine translation setup where the model is trained with source-text image pairs and tested with only source- text inputs.
Outcome:	The proposed model outperforms the baseline model on the task and setup, helping yield translations with better completeness, relevance and fluency without relying on paired images.

Copied to clipboard

Xin Zhang, Yong Jiang, Xiaobin Wang, Xuming Hu, Yueheng Sun, Pengjun Xie, Meishan Zhang

XNLP: An Interactive Demonstration System for Universal Structured NLP (2024.acl-demos)

Challenge:	Successful Named Entity Recognition models fail on texts from some special domains, for example, Chinese addresses and e-commerce titles.
Approach:	They propose to enhance NER models with correlated samples to help the text understanding . they draw correlated texts by the sparse BM25 retriever from large-scale in-domain unlabeled data .
Outcome:	Empirical results show that NER models can be enhanced with correlated samples . the proposed model can be used to reason out the correct answer on hard cases .

Copied to clipboard

Hao Fei, Meishan Zhang, Min Zhang, Tat-Seng Chua

Challenge:	Structured Natural Language Processing (XNLP) is an important subset of NLP that entails understanding the underlying semantic or syntactic structure of texts.
Approach:	They propose a XNLP demonstration system that leverages LLM to achieve universal XnLP with one model for all with high generalizability.
Outcome:	The proposed system advances in multiple aspects, including universal XNLP modeling, high performance, interpretability, scalability, and interactivity.

Crowdsourcing Learning as Domain Adaptation: A Case Study on Named Entity Recognition (2021.acl-long)

Copied to clipboard

Xin Zhang, Guangwei Xu, Yueheng Sun, Meishan Zhang, Pengjun Xie

Low-Resource NLP

Challenge:	Experimental results show that crowdsourced annotations are highly effective under supervised conditions.
Approach:	They propose an annotator-aware representation learning model that is inspired by domain adaptation methods which attempt to capture effective domain-alike features.
Outcome:	The proposed model is highly effective on a benchmark dataset and achieves state-of-the-art performance with only a very small scale of expert annotations.

BrainECHO: Semantic Brain Signal Decoding through Vector-Quantized Spectrogram Reconstruction for Whisper-Enhanced Text Generation (2025.findings-acl)

Copied to clipboard

Jilong Li, Zhenxi Song, Jiaqi Wang, Meishan Zhang, Honghai Liu, Min Zhang, Zhiguo Zhang

A Two-Stage Adaptation of Large Language Models for Text Ranking (2024.findings-acl)

Challenge:	Current EEG/MEG-to-text decoding systems rely on teacher-forcing methods . pre-trained large language models are over-dominant in decoding text from brain activity .
Approach:	They propose a framework that employs decoupled representation learning to achieve state-of-the-art performance on EEG and MEG datasets.
Outcome:	The proposed framework achieves state-of-the-art performance on EEG and MEG datasets.

Copied to clipboard

Longhui Zhang, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Meishan Zhang, Min Zhang

Constructing Code-mixed Universal Dependency Forest for Unbiased Cross-lingual Relation Extraction (2023.findings-acl)

Challenge:	Recent advances in pre-trained language models (PLMs) have significantly improved ranking performance in text ranking tasks.
Approach:	They propose a two-stage progressive paradigm to better adapt LLMs to text ranking by conducting continual pre-training on a large weakly-supervised corpus and performing SFT on high-quality data.
Outcome:	The proposed approach outperforms previous methods on in- and out-domain scenarios.

Copied to clipboard

Hao Fei, Meishan Zhang, Min Zhang, Tat-Seng Chua

Adaptive Detoxification: Safeguarding General Capabilities of LLMs through Toxicity-Aware Knowledge Editing (2025.findings-acl)

Challenge:	Recent efforts on cross-lingual relation extraction (XRE) leverage language-consistent structural features from the universal dependency resource.
Approach:	They propose to construct a type of code-mixed UD forest that combines UD and source-/target-side UD structures to achieve unbiased transfer.
Outcome:	The proposed UD forest achieves significant performance gains on ACE XRE benchmark datasets.

Copied to clipboard

Yifan Lu, Jing Li, Yigeng Zhou, Yihui Zhang, Wenya Wang, Xiucheng Li, Meishan Zhang, Fangming Liu, Jun Yu, Min Zhang

Challenge:	Existing knowledge editing methods for large language models (LLMs) suffer from over-editing, where detoxified models reject legitimate queries, compromising overall performance.
Approach:	They propose a toxicity-aware knowledge editing approach that dynamically detects toxic activation patterns during forward propagation and then routes computations through adaptive inter-layer pathways to mitigate toxicity effectively.
Outcome:	The proposed method outperforms existing methods on large language models and enhances the SafeEdit benchmark.

Cross-Lingual Dependency Parsing Using Code-Mixed TreeBank (D19-1)

Copied to clipboard

Meishan Zhang, Yue Zhang, Guohong Fu

DRTS Parsing with Structure-Aware Encoding and Decoding (2020.acl-main)

Challenge:	Treebank translation is a promising method for cross-lingual transfer of syntactic dependency knowledge.
Approach:	They propose to map dependency arcs from source treebank to target translation according to word alignments.
Outcome:	Experiments on university dependency treebanks show that translated treebank translations are more effective than translated treebans.

Copied to clipboard

Qiankun Fu, Yue Zhang, Jiangming Liu, Meishan Zhang

A Pilot Study on Dialogue-Level Dependency Parsing for Chinese (2023.findings-acl)

Challenge:	Discourse representation tree structure (DRTS) parsing is a new semantic parser which ignores structural information.
Approach:	They propose a structural-aware model to integrate structural information into the model . they use graph attention network (GAT) to exploit structural information for effective modeling .
Outcome:	The proposed model can achieve the best performance on a benchmark dataset.

Copied to clipboard

Gongyao Jiang, Shuang Liu, Meishan Zhang, Min Zhang

Syntactic Text Processing

Challenge:	Dialogue-level dependency parsing has received insufficient attention, especially for Chinese.
Approach:	They propose a signal-based method to transform seen syntactic dependencies into unseen ones between elementary discourse units (EDUs) they apply single-view and multi-view data selection to access reliable pseudo-labeled instances.
Outcome:	The proposed method transforms seen syntactic dependencies into unseen ones between elementary discourse units (EDUs) the proposed method also provides reliable pseudo-labeled instances.

Chinese Opinion Role Labeling with Corpus Translation: A Pivot Study (2021.emnlp-main)

Copied to clipboard

Ranran Zhen, Rui Wang, Guohong Fu, Chengguo Lv, Meishan Zhang

A Fine-Grained Domain Adaption Model for Joint Word Segmentation and POS Tagging (2021.emnlp-main)

Challenge:	Unlike most of the previous work focusing on the English language, this paper focuses on the Chinese ORL task.
Approach:	They propose to use a standard English MPQA dataset to construct a Chinese ORL dataset and investigate the effectiveness of cross-lingual transfer methods.
Outcome:	The proposed method is able to detect and improve the performance of the proposed method in Chinese.

Copied to clipboard

Peijie Jiang, Dingkun Long, Yueheng Sun, Meishan Zhang, Guangwei Xu, Pengjun Xie

Syntactic Text Processing

Challenge:	Experimental results show that joint models of word segmentation and POS tagging can lead to better performance because they are closely related.
Approach:	They propose a domain adaption method for Chinese word segmentation and POS tagging that uses a simple metric to model the gaps between target and target domains.
Outcome:	The proposed method can gain significant performance improvements over baselines on a benchmark dataset.

Identifying Chinese Opinion Expressions with Extremely-Noisy Crowdsourcing Annotations (2022.acl-long)

Copied to clipboard

Xin Zhang, Guangwei Xu, Yueheng Sun, Meishan Zhang, Xiaobin Wang, Min Zhang

Sentiment Analysis

Challenge:	Recent works of opinion expression identification (OEI) rely heavily on the quality and scale of the manually-constructed training corpus.
Approach:	They propose to use crowdsourcing annotations to build a large-scale but quality-unguaranteed corpus for opinion expression identification in Chinese.
Outcome:	The proposed model can be trained with a synthetic expert and is highly consistent with the training and testing phase.

A Graph-Based Neural Model for End-to-End Frame Semantic Parsing (2021.emnlp-main)

Copied to clipboard

ZhiChao Lin, Yueheng Sun, Meishan Zhang

Robust Self-Augmentation for Named Entity Recognition with Meta Reweighting (2022.naacl-main)

Challenge:	Existing studies focus on frame semantic parsing as a graph construction problem.
Approach:	They propose an end-to-end neural model to tackle frame semantic parsing jointly.
Outcome:	The proposed model is highly competitive and performs better than pipeline models on two benchmark datasets.

Copied to clipboard

Linzhi Wu, Pengjun Xie, Jie Zhou, Meishan Zhang, Ma Chunping, Guangwei Xu, Min Zhang

Robustness in NLP

Challenge:	Prior research has focused on reducing noise for specific methods to achieve an effective integration.
Approach:	They propose to use token substitution and mixup to improve named entity recognition (NER) using a meta-reweighting strategy, which is extensible and requires little effort.
Outcome:	The proposed method is extensible, imposing little effort on a specific self-augmentation method.

Unsupervised Boundary-Aware Language Model Pretraining for Chinese Sequence Labeling (2022.emnlp-main)

Copied to clipboard

Peijie Jiang, Dingkun Long, Yanzhao Zhang, Pengjun Xie, Meishan Zhang, Min Zhang

End to End Chinese Lexical Fusion Recognition with Sememe Knowledge (2020.coling-main)

Challenge:	Experimental results show that Boundary-Aware BERT can improve Chinese sequence labeling tasks.
Approach:	They propose to encode boundary information directly into pre-trained language models . they propose to use unsupervised boundary information instead of supervised boundary info .
Outcome:	The proposed architecture improves Chinese sequence labeling tasks on ten benchmarks.

Copied to clipboard

Yijiang Liu, Meishan Zhang, Donghong Ji

Recognizing Everything from All Modalities at Once: Grounded Multimodal Universal Information Extraction (2024.findings-acl)

Challenge:	a new task for coreference recognition is presented in linguistics . the fusion word is always out-of-vocabulary (OOV) words in downstream paragraph-level tasks .
Approach:	They propose a Chinese lexical fusion recognition task which could be regarded as one kind of coreference recognition.
Outcome:	The proposed model is effective and competitive for the proposed task.

Copied to clipboard

Meishan Zhang, Hao Fei, Bin Wang, Shengqiong Wu, Yixin Cao, Fei Li, Min Zhang

HiTrans: A Transformer-Based Context- and Speaker-Sensitive Model for Emotion Detection in Conversations (2020.coling-main)

Challenge:	Existing studies on IE tasks have focused on recognizing and analyzing cross-modal information . a multimodal large language model (MLLM) is developed to analyze IE across modalities .
Approach:	They propose a multimodal large language model (MLLM) capable of grounding information from all modalities.
Outcome:	The proposed framework provides a framework to analyze IE tasks over various modalities and their fine-grained groundings.

Copied to clipboard

Jingye Li, Donghong Ji, Fei Li, Meishan Zhang, Yijiang Liu

Dialogue Systems & Conversational Agents

Challenge:	Emotion detection in conversations is to detect the emotion for each utterance in conversations that have multiple speakers.
Approach:	They propose a transformer-based context- and speaker-sensitive model for EDC . they utilize a low-level transformer to generate local utterance representations .
Outcome:	The proposed model outperforms state-of-the-art models on three benchmark datasets.

Dynamic Long Context Reasoning over Compressed Memory via End-to-End Reinforcement Learning (2026.acl-long)

Copied to clipboard

Zhuoen Chen, Dongfang Li, Meishan Zhang, Baotian Hu, Min Zhang

Responsible & Trustworthy NLP

Challenge:	Large Language Models face quadratic computational costs, information forgetting, and context fragmentation . current methodologies diverge into three paradigms, each facing inherent trade-offs between Equal contribution.
Approach:	They propose a framework that enables efficient long-context inference via chunk-wise compression and selective memory recall.
Outcome:	The proposed framework reduces peak GPU memory usage and speeds up inference on multi-hop reasoning benchmarks.

PromptST: Abstract Prompt Learning for End-to-End Speech Translation (2023.emnlp-main)

Copied to clipboard

Tengfei Yu, Liang Ding, Xuebo Liu, Kehai Chen, Meishan Zhang, Dacheng Tao, Min Zhang

Towards Demonstration-Aware Large Language Models for Machine Translation (2024.findings-acl)

Challenge:	Experimental results show that PromptST can improve speech-to-text translation by capturing richer linguistic knowledge.
Approach:	They propose a plug-in prompt-enhanced S2T model that captures richer linguistic knowledge . they use a 10GB linguistic probing benchmark to investigate the fusion of speech and text features .
Outcome:	The proposed model can improve on a strong baseline by capturing richer linguistic knowledge.

Copied to clipboard

Chen Li, Meishan Zhang, Xuebo Liu, Zhaocong Li, Derek Wong, Min Zhang

ERCThinker: Fast-Slow Thinking for Emotion Recognition in Conversation (2026.acl-long)

Challenge:	Large language models for machine translation often face difficulties in leveraging demonstrations to further improve their performance.
Approach:	They propose a novel approach that integrates demonstration-aware training and inference strategies within the framework of tuning-based LTMs.
Outcome:	The proposed model integrates demonstration-aware training and inference strategies within tuning-based LTMs.

Copied to clipboard

Yumeng Fu, Weitao Huang, Junjie Wu, Hao Teng, Shouduo Shang, Meishan Zhang, Bingquan Liu

Reasoning

Challenge:	Existing methods for ERC lack interpretability and shallow semantics capture deep semantics.
Approach:	They propose a Fast-Slow thinking framework for Emotion Recognition in Conversation . they use fine-grained emotion reasoning chains to capture deep semantics .
Outcome:	The proposed framework achieves state-of-the-art in explanation and judgment on a benchmark dataset.

Enhancing Opinion Role Labeling with Semantic-Aware Word Representations from Semantic Role Labeling (N19-1)

Copied to clipboard

Meishan Zhang, Peili Liang, Guohong Fu

On the Role of Pre-trained Language Models in Word Ordering: A Case Study with BART (2022.coling-1)

Challenge:	Existing work on opinion role labeling (ORL) is highly correlative with semantic role labeled (SRL) SRL is used to identify opinion holders and holder expressions for a given predicate.
Approach:	They propose a method to enhance opinion role labeling by presenting semantic-aware word representations which are learned from SRL.
Outcome:	The proposed method outperforms two other methods on a benchmark MPQA corpus and achieves higher F scores.

Copied to clipboard

Zebin Ou, Meishan Zhang, Yue Zhang

ToolOmni: Enabling Open-World Tool Use via Agentic learning with Proactive Retrieval and Grounded Execution (2026.acl-long)

Challenge:	Existing work uses linear models and neural networks for word ordering, yet pre-trained language models have not been studied in word ordering.
Approach:	They propose a constrained language generation task using unordered words as input.
Outcome:	The proposed model is able to perform better than existing models and proves to be reliable.

Copied to clipboard

Shouzheng Huang, Meishan Zhang, Baotian Hu, Min Zhang

Information Retrieval

Challenge:	Existing methods for large language models struggle to align user intent with tool semantics or generalize to unseen tools.
Approach:	They propose a framework that enables LLMs for open-world tool use by proactive retrieval and grounded execution within a reasoning loop.
Outcome:	The proposed framework surpasses baselines in retrieval and execution efficiency by +10.8%.

Cross-Lingual Semantic Role Labeling with High-Quality Translated Training Corpus (2020.acl-main)

Copied to clipboard

Hao Fei, Meishan Zhang, Donghong Ji

A Span-Based Model for Joint Overlapped and Discontinuous Named Entity Recognition (2021.acl-long)

Challenge:	Existing approaches to semantic role labeling (SRL) are focusing on the English language.
Approach:	They propose a method for semantic role labeling that uses corpus translation to build training datasets from SRL annotations.
Outcome:	The proposed method is highly effective and can improve the target-language performance significantly.

Copied to clipboard

Fei Li, ZhiChao Lin, Meishan Zhang, Donghong Ji

Syntax-Enhanced Neural Machine Translation with Syntax-Aware Word Representations (N19-1)

Challenge:	Existing models for named entity recognition (NER) focus on overlapped or discontinuous entities.
Approach:	They propose a span-based named entity recognition model that can recognize both overlapped and discontinuous entities jointly.
Outcome:	The proposed model can recognize overlapped and discontinuous entities jointly.

Copied to clipboard

Meishan Zhang, Zhenghua Li, Guohong Fu, Min Zhang

APGN: Adversarial and Parameter Generation Networks for Multi-Source Cross-Domain Dependency Parsing (2021.findings-emnlp)

Challenge:	Syntax integration has been demonstrated highly effective in neural machine translation (NMT).
Approach:	They propose a method to integrate source-side syntax implicitly for neural machine translation . they use hidden representations of a well-trained end-to-end dependency parser to concatenate them with ordinary word embeddings to enhance basic NMT models.
Outcome:	The proposed method outperforms existing methods on two translation tasks . it can be easily integrated into the widely-used sequence-to-sequence (Seq2Sequen) framework .

Copied to clipboard

Ying Li, Meishan Zhang, Zhenghua Li, Min Zhang, Zhefeng Wang, Baoxing Huai, Nicholas Jing Yuan

Robustness in NLP

Challenge:	Existing models for dependency parsing use labeled training data for several fixed domains, but performance drops when labeles only exist for several out-domains.
Approach:	They propose a model for multi-source cross-domain dependency parsing that uses a parameter generation network and adversarial network for learning domain-invariant representations.
Outcome:	The proposed model improves cross-domain parsing performance by about 2 points over strong BERT-enhanced baselines over a recently released dataset for multi-domain dependency parse.

LaERC-S: Improving LLM-based Emotion Recognition in Conversation with Speaker Characteristics (2025.coling-main)

Copied to clipboard

Yumeng Fu, Junjie Wu, Zhongjie Wang, Meishan Zhang, Lili Shan, Yulin Wu, Bingquan Liu

Sentiment Analysis

Challenge:	Emotion recognition in conversation (ERC) is a task of discerning human emotions for each utterance within a conversation.
Approach:	They propose a framework that uses large language models to analyze speaker characteristics . they use two-stage learning to make the models reason speaker characteristics and track emotion of the speaker .
Outcome:	The proposed framework outperforms existing methods on three benchmark datasets.

Learning to Extract Rational Evidence via Reinforcement Learning for Retrieval-Augmented Generation (2026.findings-acl)

Copied to clipboard

Xinping Zhao, Shouzheng Huang, Yan Zhong, Xinshuo Hu, Meishan Zhang, Baotian Hu, Min Zhang

Reasoning

Challenge:	Retrieval-Augmented Generation (RAG) is effective in Large Language Models (LLMs). However, retrieval noises undermine the quality of LLMs’ generation, necessitating the development of denoising mechanisms.
Approach:	They propose a model which integrates reasoning and extracting into one unified trajectory, followed by knowledge token masking to avoid information leakage.
Outcome:	Extensive experiments on five benchmark datasets show the superiority of EviOmni, which provides compact and high-quality evidence, enhances the accuracy of downstream tasks, and supports both traditional and agentic RAG systems.

JoPR: Joint Emotion Perception and Reasoning for Conversational Emotion Recognition (2026.acl-long)

Copied to clipboard

Yumeng Fu, Weitao Huang, Junjie Wu, Hao Teng, Meishan Zhang, Bingquan Liu

Dialogue Systems & Conversational Agents

Challenge:	Existing methods for ERC lack human-like emotion reasoning and discrimination between similar emotions.
Approach:	They propose a multi-dimension curriculum with long CoT fine-tuning to clone human-like emotion reasoning for conversational emotion recognition.
Outcome:	The proposed model outperforms existing methods on three widely used datasets and shows that it is more intuitive and more accurate.

Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment (2026.acl-long)

Copied to clipboard

Bobo Li, Wu Rui, Zibo Ji, Meishan Zhang, Hao Fei, Min Zhang, Mong-Li Lee, Wynne Hsu

Reasoning

Challenge:	Large Language Models (LLMs) have enabled the development of powerful autonomous systems.
Approach:	They propose a model trained through dialectical alignment to enforce perspective-invariant reasoning.
Outcome:	The proposed model mitigates attribution inconsistency and significantly improves fault resolution rates in ambiguous scenarios.

Contrastive Learning on LLM Back Generation Treebank for Cross-domain Constituency Parsing (2025.acl-long)

Copied to clipboard

Peiming Guo, Meishan Zhang, Jianling Li, Min Zhang, Yue Zhang

LLMs Can Also Do Well! Breaking Barriers in Semantic Role Labeling via Large Language Models (2025.findings-acl)

Challenge:	Existing constituency treebanks are limited in out-of-domain settings, therefore constituency parsing is still a challenge.
Approach:	They propose a novel method for constituency parsing using large language models . they use a cross-domain constituency treebank to fill missing words with the incomplete one .
Outcome:	The proposed method achieves state-of-the-art performance on average compared with baselines on five target domains of MCTB.

Copied to clipboard

Xinxin Li, Huiyao Chen, Chengjun Liu, Jing Li, Meishan Zhang, Jun Yu, Min Zhang

Generating Visual Spatial Description via Holistic 3D Scene Understanding (2023.acl-long)

Challenge:	Semantic role labeling (SRL) is a crucial task of natural language processing (NLP).
Approach:	They propose to equip LLMs with retrieval-augmented generation and self-correction mechanisms to enable SRL to perform better in Chinese and English.
Outcome:	The proposed method achieves state-of-the-art in Chinese and English on three widely-used benchmarks.

Copied to clipboard

Yu Zhao, Hao Fei, Wei Ji, Jianguo Wei, Meishan Zhang, Min Zhang, Tat-Seng Chua

Speed Up Your Code: Progressive Code Acceleration Through Bidirectional Tree Editing (2025.acl-long)

Challenge:	Existing VSD work focuses on skewed spatial understanding of target objects . Existing work merely models the 2D geometrical vision features .
Approach:	They propose to incorporate 3D scene features into visual spatial description tasks by sampling topologically-diverse subgraphs from Go3D-S2G.
Outcome:	The proposed framework outperforms baselines on two VSD datasets and produces more spatially-diversified generation.

Copied to clipboard

Longhui Zhang, Jiahao Wang, Meishan Zhang, GaoXiong Cao, Ensheng Shi, Mayuchi Mayuchi, Jun Yu, Honghai Liu, Jing Li, Min Zhang

Programming Languages in NLP

Challenge:	Existing training methods, such as direct instruction fine-tuning, overlook hierarchical relationships among acceleration patterns.
Approach:	They propose a new training paradigm that uses bidirectional tree editing and progressive code acceleration learning to improve LLMs’ CA capabilities.
Outcome:	The proposed training paradigm outperforms prompt-enhanced GPT-4 and current training-based methods on average across five programming languages.

Less Languages, Less Tokens: An Efficient Unified Logic Cross-lingual Chain-of-Thought Reasoning Framework (2026.acl-long)

Copied to clipboard

Chenyuan Zhang, Qiguang Chen, Xie Chen, Zhuotao Tian, Bowen Xing, Meishan Zhang, Libo Qin, Baotian Hu, Min Zhang

RST Discourse Parsing with Second-Stage EDU-Level Pre-training (2022.acl-long)

Challenge:	Existing methods for cross-lingual chain-of-thought (XCoT) with self-consistency are costly due to extensive sampling of full trajectories across languages.
Approach:	They propose a cross-lingual chain-of-thought framework that minimizes redundancy in token usage and latency.
Outcome:	Experiments on polymath show that UL-XCoT reduces decoding token costs and latency by 50% . UL XCot also aggregates remaining high-quality reasoning paths via voting .

Copied to clipboard

Nan Yu, Meishan Zhang, Guohong Fu, Min Zhang

Towards Text-Image Interleaved Retrieval (2025.acl-long)

Challenge:	Existing pre-trained language models (PLMs) are based on sentence-level pre-training, which is different from the basic processing unit, i.e. element discourse unit (EDU).
Approach:	They propose a second-stage EDU-level pre-training approach to learn effective EDU representations continually based on well pre-trained language models.
Outcome:	The proposed method improves F1 score by 2.1 points on a benckmark dataset.

Copied to clipboard

Xin Zhang, Ziqi Dai, Yongqi Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Meishan Zhang, Jun Yu, Wenjie Li, Min Zhang

LLM-enhanced Self-training for Cross-domain Constituency Parsing (2023.emnlp-main)

Challenge:	Existing multimodal information retrieval models rely on single-image inputs . current models use a dense retrieval paradigm, but this approach is not effective .
Approach:	They propose a text-image interleaved retrieval task where query and document are interleaves . they adapt off-the-shelf retrievers and build a dense baseline by interleaded multimodal large language model .
Outcome:	The proposed model achieves significant improvements over the baseline by substantially fewer visual tokens.

Copied to clipboard

Jianling Li, Meishan Zhang, Peiming Guo, Min Zhang, Yue Zhang