Boyan Xu

Boyan Xu

Guangdong University of Technology

RESEARCHER · GDUT

Boyan Xu 许柏炎

School of Computer Science · Guangdong University of Technology

I am a researcher at Guangdong University of Technology (GDUT), working with Prof. Ruichu Cai in the Data Mining & Information Retrieval Laboratory. My research focuses on Text-to-SQL, Large Language Models, and Causality-enhanced AI, with additional interests in affective computing and natural language processing.

广东工业大学计算机学院硕士生导师，研究方向包括大语言模型、Text-to-SQL 与因果增强智能。

Google Scholar ACL Anthology GDUT Profile

My research connects natural language, structured data, and causal reasoning:

  • Text-to-SQL & Data Agents: Natural language interfaces for databases, schema linking, and multi-turn reasoning.
  • Large Language Models: Retrieval-augmented generation, structured reasoning, and language understanding.
  • Causality-enhanced AI: Causal discovery foundation models and their connections to data analysis and intelligent decision-making.

I welcome collaboration with motivated students and industry partners on Text-to-SQL agents, causality-enhanced AI, and ABSA. Feel free to email me with your CV and a short statement of interest.

🔥 News

  • 2026.07:  CDFM, a causal discovery foundation model for zero-shot structural inference, is now available. Paper · Model weights.
  • 2026.04:  🎉🎉 4 papers(2 main and 2 findings) under my supervision have been accepted by ACL 2026.
  • 2025.12:  🎉🎉 We introduce a new benchmark for Indonesian Multimodal Emotion Recognition, accompanying OmniMER.
  • 2025.08:  🎉🎉 GenLink have been accepted by EMNLP 2025 Main.
  • 2025.01:  🎉🎉 3 papers under my supervision have been accepted by NAACL 2025.
  • 2025.01:  🎉🎉 Chat2DB have been accepted by ICDE 2025 Demo track.
  • 2024.11:  🎉🎉 2 papers under my supervision have been accepted by COLING 2025.
  • 2024.07:  🎉🎉 I am the problem setter for the 2024 Third International Algorithm Case Competition (IACC), hosted by Pazhou Lab. The challenge I designed focuses on “Generating Database Query Commands Based on Large Language Models”.
  • 2024.05:  🎉🎉 The paper I supervised, “S2GSL: Incorporating Segment to Syntactic Enhanced Graph Structure Learning for Aspect-based Sentiment Analysis” has been accepted by ACL 2024 Main.
  • 2023.04:  🎉🎉 My leader launched Chat2DB, a conversational AI product designed to access private databases or tabular data. With Chat2DB, users don’t need to learn technical principles or use specialized tools. By simply uploading their data or database and describing their requirements in the chatbox, they can receive results within seconds.

📝 Publications

†Corresponding Author, *Equal Contribution

Preprint 2026
CDFM overview: causal knowledge guides synthetic data generation and foundation-model pretraining for causal discovery

CDFM: Towards a General-Purpose Causal Discovery Foundation Model

Jie Qiao, Ruichu Cai, Zijian Li, Weilin Chen, Pengfei Hua, Boyan Xu, Zhengming Chen, Zhifeng Hao, Peng Cui

Code · Model weights

  • Pretrained on synthetic structural causal models, CDFM infers causal graphs from observational data in a single forward pass for zero-shot causal discovery.
ACL 2026(Main)
sym

MMErroR: A Benchmark for Erroneous Reasoning in Vision-Language Models

Yang Shi, Yifeng Xie, Minzhe Guo, Liangsi Lu, Mingxuan Huang, Jingchao Wang, Zhihong Zhu, Boyan Xu†, Zhiqi Huang

Code · Project · Dataset

  • We introduce MMErroR, a multi-modal benchmark for process-level error diagnosis in vision-language models, comprising 1,997 curated samples across 24 subdomains and four reasoning error types, revealing that even the strongest evaluated VLM achieves only 66.65% accuracy in fine-grained error-type classification.
ACL 2026(Main)
sym

Rose-SQL: Role-State Evolution Guided Structured Reasoning for Multi-Turn Text-to-SQL

Le Zhou, Feng Yao, Fengcai Qiao, Bo Xu, Fangyuan Wang, Boyan Xu†

  • We propose Rose-SQL, a training-free framework that adapts small-scale reasoning models to multi-turn Text-to-SQL through Role-State evolution, using gain dependency analysis, structural isomorphism checks, and augmented hierarchical reasoning to track dialogue changes and guide SQL generation, achieving state-of-the-art performance on SParC and CoSQL without task-specific fine-tuning.
ACL 2026(Findings)
sym

SAM-NER: Semantic Archetype Mediation for Zero-Shot Named Entity Recognition

Ruichu Cai, Juntao Gan, Miao Mai, Zhifeng Hao, Boyan Xu†

  • We propose SAM-NER, a three-stage zero-shot NER framework that mitigates semantic drift through semantic archetype mediation, combining cooperative entity discovery, universal archetype projection, and definition-guided calibration to achieve state-of-the-art cross-domain performance on CrossNER without relying on external knowledge bases.
ACL 2026(Findings)
sym

SERE: Structural Example Retrieval for Enhancing LLMs in Event Causality Identification

Zhifeng Hao, Zhongjie Chen, Junhao Lu, Shengyin Yu, Guimin Hu, Keli Zhang, Ruichu Cai, Boyan Xu†

Code

  • We propose SERE, a structural example retrieval framework that enhances LLMs for event causality identification by jointly leveraging conceptual path similarity, syntactic tree similarity, and causal pattern filtering to retrieve structurally aligned demonstrations, mitigating causal hallucination and improving performance across multiple ECI datasets.
EMNLP 2025(Main)
sym

GenLink: Generation-Driven Schema-Linking via Multi-Model Learning for Text-to-SQL

Zhifeng Hao, Junqi Huang, Shaobin Shi, Ruichu Cai, Boyan Xu†

Code

  • We propose GenLink, a generation-driven schema-linking framework that integrates multiple small language models to infer implicit schema links through SQL generation and self-consistency, achieving strong cross-domain Text-to-SQL performance on BIRD and Spider with execution accuracies of 67.34%, 89.7%, and 87.8%.
ICDE 2025(Demo)
sym

Chat2DB: Chatting to the Database with Interactive Agent Assisted Language Models

Boyan Xu *, Yuyuan Cai *, Shaobin Shi, Zhifeng Hao, Ruichu Cai

  • We propose Chat2DB, an interactive conversational database system that enhances Text-to-SQL parsers through agent-assisted question-schema linking and adaptive retraining, enabling users to collaboratively refine schema selection, improve SQL generation accuracy, and query domain-specific databases through natural language.
NAACL 2025(Main)
sym

Handling Missing Entities in Zero-Shot Named Entity Recognition: Integrated Recall and Retrieval Augmentation

Ruichu Cai, Junhao Lu, Zhongjie Chen, Boyan Xu†, Zhifeng Hao

Code

  • We introduce IRRA, a novel two-stage framework that first uses recall-augmented entity extraction on perturbed data to boost recall and then applies retrieval-augmented generation for type correction, achieving state-of-the-art zero-shot cross-domain NER performance.
NAACL 2025(Main)
sym

Track-SQL: Enhancing Generative Language Models with Dual-Extractive Modules for Schema and Context Tracking in Multi-turn Text-to-SQL

Bingfeng chen, Shaobin Shi, yongqi luo, Boyan Xu†, Ruichu Cai, Zhifeng Hao

Code

  • We propose Track-SQL, a dual-extractive framework that augments generative LMs with a semantic-enhanced schema extractor and a schema-aware context extractor to track schema and context changes in multi-turn Text-to-SQL, achieving state-of-the-art results on SparC and CoSQL with execution accuracy gains of 7.1% and 9.55%.
NAACL 2025(Findings)
sym

S^2IT: Stepwise Syntax Integration Tuning for Large Language Models in Aspect Sentiment Quad Prediction

Bingfeng chen, Chenjie Qiu, Yifeng Xie, Boyan Xu†, Ruichu Cai, Zhifeng Hao

Code

  • We propose S^2IT, a novel Stepwise Syntax Integration Tuning framework that progressively incorporates global and local syntactic structure knowledge into LLMs to significantly enhance performance on Aspect Sentiment Quad Prediction.
COLING 2025(Oral)
sym

Dr.ECI: Infusing Large Language Models with Causal Knowledge for Decomposed Reasoning in Event Causality Identification

Ruichu Cai, Shengyin Yu, Jiahao Zhang, Wei Chen, Boyan Xu†, Keli Zhang

Code

  • We propose Dr.ECI, a multi-agent decomposed reasoning framework for Event Causality Identification that employs specialized discovery agents (Causal Explorer, Mediator Detector) and reasoning agents (Direct and Indirect Reasoners) to capture implicit, indirect, and generalized causal structures, achieving state-of-the-art performance.
COLING 2025(Oral)
sym

CACA: Context-Aware Cross-Attention Network for Extractive Aspect Sentiment Quad Prediction

Bingfeng Chen, Haoran Xu, Yongqi Luo, Boyan Xu†, Ruichu Cai, Zhifeng Hao

Code

  • We propose CACA, an extractive ASQP framework that employs a Context-Aware Cross-Attention Network—alternating updates of explicit and implicit representations—and contrastive learning to align aspects and opinions for implicit term prediction, achieving superior results on three benchmarks.
ACL 2024(Main)
sym

S^2GSL: Incorporating Segment to Syntactic Enhanced Graph Structure Learning for Aspect-based Sentiment Analysis

Bingfeng Chen, Qihan Ouyang, Yongqi Luo, Boyan Xu†, Ruichu Cai, Zhifeng Hao (ACL 2024 Main)

Code

  • We introduce S^2GSL, a novel graph-structure learning framework for ABSA that combines segment-aware semantic graph learning with syntax-based latent graph learning—and a self-adaptive aggregation network to filter irrelevant contexts and fuse diverse structures—achieving superior results on four benchmarks.