Haoxuan Zhang
Ph.D. student in Information Science (Data Science)
IDEA Lab, University of North Texas, Denton, Texas
I study how scientific knowledge is produced and evaluated in the AI era.
Research output increasingly takes the form of AI models and datasets, and large language models increasingly write the summaries and reviews that science relies on. My work addresses two questions this raises. AI documentation asks how this knowledge is recorded: I build benchmarks and methods for generating model and data cards that state what a model or dataset is, how it was built, and where it falls short. Knowledge evaluation asks how its quality is judged: I measure novelty and breakthrough innovation in science and test whether LLM-generated reviews and summaries can be trusted. The two depend on each other. Without documentation, AI work cannot be evaluated, compared, or reused; and documentation written by LLMs is itself knowledge that has to be evaluated.
I am a member of the Intelligent Data Engineering and Analytics (IDEA) Lab, advised by Prof. Haihua Chen, Prof. Ting Xiao, and Prof. Yang Zhang.
News
Latest: TopoGraphRAG-Bench accepted to NeurIPS 2026
TopoGraphRAG-Bench was accepted to the NeurIPS 2026 Evaluations and Datasets Track.
Two papers accepted to Findings of EMNLP 2026: MetaGAI and a survey of human–agent collaboration in scientific innovation.
LongSumEval accepted to IEEE AITest 2026, and LLM-ReSum to IEEE BigDataService 2026.
AdaQE-CG accepted to WWW 2026.
Two papers accepted: ReviewGuard at JCDL 2025 and our study of LLM-generated peer reviews at ICDM 2025, which went on to receive the Best Poster Paper award.
Started my Ph.D. at the University of North Texas.
Publications
Knowledge production and evaluation in the AI era
-
TopoGraphRAG-Bench: Evaluating multimodal GraphRAG on layout-grounded evidence reasoning
In Advances in Neural Information Processing Systems (NeurIPS 2026), Evaluations and Datasets Track. OpenReview
@inproceedings{li2026topographrag, title = {{TopoGraphRAG-Bench}: Evaluating Multimodal {GraphRAG} on Layout-Grounded Evidence Reasoning}, author = {Li, R. and Lin, J. and Zhang, H. and Chen, H. and Ding, J. and Gehringer, E. and Zhang, Y.}, booktitle = {Advances in Neural Information Processing Systems (NeurIPS), Evaluations and Datasets Track}, year = {2026}, url = {https://openreview.net/forum?id=pXhvOZB8Qw} } -
MetaGAI: A large-scale and high-quality benchmark for generative AI model and data card generation
In Findings of the Association for Computational Linguistics: EMNLP 2026 (pp. 34910–34931). ACL Anthology arXiv
@inproceedings{zhang2026metagai, title = {{MetaGAI}: A Large-Scale and High-Quality Benchmark for Generative {AI} Model and Data Card Generation}, author = {Zhang, H. and Li, R. and Zhang, Y. and Liang, Z. and Ding, J. and Xiao, T. and Chen, H.}, booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2026}, pages = {34910--34931}, year = {2026}, url = {https://aclanthology.org/2026.findings-emnlp.1641/} } -
Puzzles to paradigms: A survey of human–agent collaboration in scientific innovation
In Findings of the Association for Computational Linguistics: EMNLP 2026 (pp. 35017–35041). ACL Anthology
@inproceedings{liang2026puzzles, title = {Puzzles to Paradigms: A Survey of Human--Agent Collaboration in Scientific Innovation}, author = {Liang, Z. and Zhang, H. and Chen, H. and Xiao, T. and Ding, J. and Zhang, Y.}, booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2026}, pages = {35017--35041}, year = {2026}, url = {https://aclanthology.org/2026.findings-emnlp.1645/} } -
AdaQE-CG: Adaptive query expansion for web-scale generative AI model and data card generation
In Proceedings of The Web Conference (WWW 2026). ACM DL
@inproceedings{zhang2026adaqecg, title = {{AdaQE-CG}: Adaptive Query Expansion for Web-Scale Generative {AI} Model and Data Card Generation}, author = {Zhang, H. and Li, R. and Liang, Z. and Sattari, M. and Vo, P. and Qu, C. and Xiao, T. and Ding, J. and Zhang, Y. and Chen, H.}, booktitle = {Proceedings of The Web Conference (WWW 2026)}, year = {2026}, doi = {10.1145/3774904.3792685} } -
LongSumEval: Question-answering based evaluation and feedback-driven refinement for long document summarization
In Proceedings of the 8th IEEE International Conference on Artificial Intelligence Testing (AITest 2026). arXiv
@inproceedings{nguyen2026longsumeval, title = {{LongSumEval}: Question-Answering Based Evaluation and Feedback-Driven Refinement for Long Document Summarization}, author = {Nguyen, H. and Zhang, H. and Zhang, Y. and Ding, J. and Chen, H.}, booktitle = {Proceedings of the 8th IEEE International Conference on Artificial Intelligence Testing (AITest 2026)}, year = {2026}, url = {https://arxiv.org/abs/2604.25130} } -
LLM-ReSum: A framework for LLM reflective summarization through self-evaluation
In Proceedings of the IEEE 12th International Conference on Big Data Computing Service and Machine Learning Applications (BigDataService 2026). DOI
@inproceedings{nguyen2026llmresum, title = {{LLM-ReSum}: A Framework for {LLM} Reflective Summarization through Self-Evaluation}, author = {Nguyen, H. and Zhang, H. and Zhang, Y. and Chen, H. and Ding, J.}, booktitle = {Proceedings of the IEEE 12th International Conference on Big Data Computing Service and Machine Learning Applications (BigDataService 2026)}, year = {2026}, doi = {10.1109/BigDataService70481.2026.00012} } -
ReviewGuard: Enhancing deficient peer review detection via LLM-driven data augmentation
In Proceedings of the ACM/IEEE Joint Conference on Digital Libraries (JCDL 2025). arXiv
@inproceedings{zhang2025reviewguard, title = {{ReviewGuard}: Enhancing Deficient Peer Review Detection via {LLM}-Driven Data Augmentation}, author = {Zhang, H. and Li, R. and Xiao, T. and Ding, J. and Chen, H.}, booktitle = {Proceedings of the ACM/IEEE Joint Conference on Digital Libraries (JCDL 2025)}, year = {2025}, url = {https://arxiv.org/abs/2510.16549} } -
Unveiling the merits and defects of LLMs in automatic review generation for scientific papers
In Proceedings of the IEEE International Conference on Data Mining (ICDM 2025). arXiv
@inproceedings{li2025unveiling, title = {Unveiling the Merits and Defects of {LLMs} in Automatic Review Generation for Scientific Papers}, author = {Li, R. and Zhang, H. and Xiao, T. and Ding, J. and Chen, H.}, booktitle = {Proceedings of the IEEE International Conference on Data Mining (ICDM 2025)}, year = {2025}, url = {https://arxiv.org/abs/2509.19326} } -
IBID-CCT: A novel model for interdisciplinary breakthrough innovation detection based on the cusp catastrophe theory
Information Processing & Management, 62(4), 104121. DOI
@article{wang2025ibidcct, title = {{IBID-CCT}: A Novel Model for Interdisciplinary Breakthrough Innovation Detection Based on the Cusp Catastrophe Theory}, author = {Wang, Z. and Wang, N. and Zhang, H. and Wang, Z. and Wang, Z. and Ding, J. and Chen, H.}, journal = {Information Processing \& Management}, volume = {62}, number = {4}, pages = {104121}, year = {2025}, doi = {10.1016/j.ipm.2025.104121} } -
The evolving role of large language models in scientific innovation: Evaluator, collaborator, and scientist
arXiv preprint arXiv:2507.11810. arXiv
@misc{zhang2025evolving, title = {The Evolving Role of Large Language Models in Scientific Innovation: Evaluator, Collaborator, and Scientist}, author = {Zhang, H. and Li, R. and Zhang, Y. and Xiao, T. and Chen, J. and Ding, J. and Chen, H.}, year = {2025}, eprint = {2507.11810}, archivePrefix = {arXiv} } -
An effective framework for measuring the novelty of scientific articles through integrated topic modeling and cloud model
Journal of Informetrics, 18(4), 101587. DOI
@article{wang2024novelty, title = {An Effective Framework for Measuring the Novelty of Scientific Articles through Integrated Topic Modeling and Cloud Model}, author = {Wang, Z. and Zhang, H. and Chen, J. and Chen, H.}, journal = {Journal of Informetrics}, volume = {18}, number = {4}, pages = {101587}, year = {2024}, doi = {10.1016/j.joi.2024.101587} } -
Content-based quality evaluation of scientific papers using coarse feature and knowledge entity network
Journal of King Saud University – Computer and Information Sciences, 36(6), 102119. DOI
@article{wang2024quality, title = {Content-Based Quality Evaluation of Scientific Papers Using Coarse Feature and Knowledge Entity Network}, author = {Wang, Z. and Zhang, H. and Chen, H. and Feng, Y. and Ding, J.}, journal = {Journal of King Saud University -- Computer and Information Sciences}, volume = {36}, number = {6}, pages = {102119}, year = {2024}, doi = {10.1016/j.jksuci.2024.102119} } -
Exploring and evaluating the index for interdisciplinary breakthrough innovation detection
The Electronic Library, 42(4), 536–552. DOI
@article{wang2024index, title = {Exploring and Evaluating the Index for Interdisciplinary Breakthrough Innovation Detection}, author = {Wang, Z. and Qiao, X. and Chen, J. and Li, L. and Zhang, H. and Ding, J. and Chen, H.}, journal = {The Electronic Library}, volume = {42}, number = {4}, pages = {536--552}, year = {2024}, doi = {10.1108/EL-06-2023-0141} }
Honors and awards
ICDM 2025 Best Poster Paper and 5 more
- 2025
Best Poster Paper, IEEE International Conference on Data Mining (ICDM 2025)
- 2025
Graduate Student Research Award, University of North Texas
- 2025
SIGIR Student Travel Grant, ACM Special Interest Group on Information Retrieval
- 2025
Graduate Student Travel Grant, University of North Texas
- 2024.06
Excellent Master’s Thesis, Central China Normal University
- 2022
Excellent Graduate Student, Central China Normal University
Education
Ph.D., University of North Texas; M.S. and B.S., Central China Normal University
-
2024–present
Ph.D. in Information Science (Data Science)
University of North Texas, Denton, USA
-
2021–2024
M.S. in Library and Information Science
Central China Normal University, Wuhan, China
-
2017–2021
B.S. in Electronic Business
Central China Normal University, Wuhan, China
Professional service
Website Chair for 6 conferences and workshops; reviewer for 7 journals
Organizing committees
- 2026
Website Chair, ACM/IEEE Joint Conference on Digital Libraries (JCDL 2026)
- 2025
Website Chair, 1st International Workshop on Data Mining and AI for Law, at ICDM 2025
- 2025
Website Chair, 1st Integrated Approaches to Testing Data-Centric AI Systems: Methods, Metrics, and Benchmarks (AIxSE 2025)
- 2024
Website Chair, 2nd Workshop on Innovation Measurement for Scientific Communication (IMSC) in the Era of Big Data, at JCDL 2024
- 2024
Website Chair, 1st Workshop on Utilizing AI/ML to Enhance Information Extraction, Organization, and Retrieval from Large-scale Archival Collections, at JCDL 2024
- 2023
Website Chair, 1st Workshop on Innovation Measurement for Scientific Communication (IMSC) in the Era of Big Data, at JCDL 2023
Journal reviewing
- Since 2026
Telematics and Informatics
Information Processing & Management
Scientometrics
- Since 2025
Humanities and Social Sciences Communications
Journal of the Association for Information Science and Technology
- Since 2024
The Electronic Library
Data Intelligence
Conference reviewing
ICDM, EMNLP, ACL, IEEE QRS, IEEE AITest, ACM/IEEE JCDL, and IEEE BigData
Teaching and mentoring
Teaching Assistant for DTSC 5502 since 2024
-
2024.08–present
Teaching Assistant, DTSC 5502 (Section 401): Principles and Techniques for Data Science
University of North Texas. Fall 2024, Spring 2025, Fall 2025, Spring 2026.
-
2025.05–2025.07
Research Assistant, UNT AI/CS Summer Research Program
University of North Texas