I am a second-year Ph.D. candidate in Computer Science at The Hong Kong Polytechnic University under the supervision of Professor Chen Jason Zhang. I also work closely with Professor Raymond Chi-Wing Wong. I previously earned my Master’s degree from The Hong Kong University of Science and Technology and my Bachelor of Science from Chongqing University.

My research centers on Natural Language Processing (NLP) for Database, with a particular focus on data-centric tasks. I am also interested in MLLM pre-training data construction, including efficient and automated data quality assessment models, data acquisition and curation, diversity assurance, noisy data cleaning, and coreset selection. In addition, I explore mathematical optimization techniques to enhance the performance and efficiency of Data Agents.

💡 I am seeking exchange or visiting research opportunities in 2027, as well as long-term research collaborators with shared interests. If you are interested in my work, please feel free to contact me via email: zhuoy.wan@connect.polyu.hk.

Natural Language Interfaces to Data

Connecting natural language with databases, focusing on data-centric tasks including Text-to-SQL, Text-to-Visualization, and Text-to-OverpassQL.

MLLM Pre-training Data Construction

Efficient data acquisition and curation, with a focus on diversity, noisy data cleaning, and coreset selection.

Data Agents & Optimization

Exploring autonomous data agents and mathematical optimization to improve their performance and efficiency.

  • 🎉 Two papers have been accepted to ICDE 2027 (CCF-A): my first-author paper, “XOver: Execution-Calibrated Exemplar Contextualization for Multilingual Text-to-OverpassQL on OpenStreetMap”, and our co-authored paper, “Text-to-RLS: Towards Constraint-Anchored Row-Level Security in Natural Language Database Interfaces”!

  • 🎉 Our co-authored paper, “Meta-Moderator: Empowering Multi-Agent Debate with Meta-Cognition”, has been accepted to Findings of EMNLP 2026!

  • 🎉 I was invited to serve as a Program Committee Member for WSDM 2027.

  • 🎉 My first-author paper, “TACO: Token-Aware Context Optimization for Structured OpenStreetMap Query Generation”, has been accepted to SIGMOD 2027 (First Round, CCF-A)!

Earlier updates
  • 🎉 Our co-authored paper, “SrDetection: A Self-Referential Framework for Data Leakage Detection in Code Large Language Models”, has been accepted to Findings of ACL 2026!

  • 🎉 I became a Ph.D. candidate at The Hong Kong Polytechnic University!

  • 🎉 Two papers have been accepted to ICDE 2026 (CCF-A): my first-author paper, “OsmT: Bridging OpenStreetMap Queries and Natural Language with Open-source Tag-aware Language Models”, and our co-authored paper, “VisPoison: An Effective Backdoor Attack Framework for Tabular Data Visualization Models”!

  • 🎉 Our co-authored paper, “Medical Image Debiasing by Learning Adaptive Agreement from a Biased Council”, has been accepted to Medical Image Analysis (JCR Q1)!

  • 🎉 My first-author paper, “DataVisT5: A Pre-trained Language Model for Jointly Understanding Text and Data Visualization”, has been accepted to ICDE 2025 (CCF-A)!

PUBLISHED & ACCEPTED

OsmT research figure
SrDetection research figure

SrDetection: A Self-Referential Framework for Data Leakage Detection in Code Large Language Models

Shuaimin Li, Liyang Fan, Zeyang Li, Zhuoyue Wan, Yufang Lin, Shiwen Ni, Feiteng Fang, Hamid Alinejad-Rokny, Yuanfeng Song, Kun Jing, Chen Jason Zhang, Min Yang

The Annual Meeting of the Association for Computational Linguistics ACL’26 · Findings

PREPRINTS

Second Prize, The Chinese Mathematics Competitions

Second-class Scholarship, Chongqing University

Advanced Individual of Scientific and Technological academic innovation, Chongqing University

Third Prize, The 8th TipDM Cup Data Mining Challenge Committee

Outstanding Student, Chongqing University

Outstanding Student, Chongqing University

Third-class Scholarship, Chongqing University

UNIVERSITY ADDRESS

The Hong Kong Polytechnic University

Hung Hom, Kowloon, Hong Kong · 香港九龍紅磡

zhuoy.wan@connect.polyu.hk

Address

Visitor Statistics

Research figure