I am a postdoctoral researcher in the Department of Computer Science and Technology, Tsinghua Univerisity, collaborated with Prof. Jie Tang. Prior to that, I received my Ph.D. degree from Tsinghua Univerisity, and my bachelor degree from Nanjing Univerisity.
My research interests generally include data mining, knowledge graph, and large language models, with an emphasis on developing scalable and lightweight data mining algorithms primarily on academic knowledge graph to accelerate academic research and scientific innovation. Recently, I am also working on LLM inference acceleration.
I am looking for self-motivated students to work with me on the area of speculative decoding and LLM agent. If interested, please drop me a message by email.
What’s New
- AMiner Rumination, an academic deep research system, has launched. Feel free to try it out and ask academic questions!
- OAG-Challenge at KDD Cup 2024 has set up long-term leaderboard. Welcome to join the competition and address the key challenges in academic graph mining.
- We maintain a paper sharing channel about large foundation models at Xiaohongshu. Feel free to subscribe!
Projects
LLM Agent for Academic Data
Large Language Model (LLM) Agents are revolutionizing the development of academic applications, making them smarter and more accessible. Our current focus is on pioneering technologies and products in areas such as AI-driven Academic Search, Intelligent Paper Reading, and AMiner Deep Research. Our goal is to create powerful agents that excel in tasks like planning, searching, summarizing, and reflecting with high levels of efficiency and effectiveness.
Name Disambiguation, a core challenge in academic systems, is now facing greater challenges with the rapid growth of papers. We develop WhoIsWho, a million-scale human-annotated author name disambiguation benchmark, decompose the AND problem into several subtasks, and develop specialized algorithms for each subtask.
WhoIsWho Benchmark: [code & data]
- From-Scratch Name Disambiguation:
- INcorrect Assignment Detection:
GuARD: a multi-modal-like multi-turn instruction tuning framework. [code & data & model]
BOND: an end-to-end method that bootstraps the global clustering and local metric learning. 
A representation learning method that incorporates both global and local information between documents. 
Paper Source Tracing (PST)
Comprehending the patterns of scientific evolution, such as the trends of topics and the flow of ideas, are critical for researchers in knowledge discovery. However, a notable gap persists between the large-scale and semantically rich citation relations and the backbone structure of scientific evolution. To bridge this gap, we construct a high-quality and ever-increasing dataset PST-Bench in computer science. Addtionally, we present a live demo to automatically trace the source of any target paper based on its full text. [code & data] [demo]
Open Academic Graph (OAG)
The overarching goal of academic data mining is to deepen our comprehension of the development, nature, and trends of science, unlocking immense value across scientific, technological, and educational domains. We release a series of meticulously curated academic benchmarks to foster the develeopment of this field.
OAG: A collection of billion-scale academic knowledge graphs. [data]
OAG-Bench: a comprehensive, multi-aspect, and fine-grained human-curated benchmark built on OAG, covering 10 tasks, 20 datasets, 70+baselines, and 120+experimental results to date. [code] [data] [data challenge]
Publications
Small Language Model Makes an Effective Long Text Extractor
The 39th Annual AAAI Conference on Artificial Intelligence (AAAI) 2025
Yelin Chen, Fanjin Zhang, Jie Tang
[pdf] [code & data]
GuARD: Effective Anomaly Detection through a Text-Rich and Graph-Informed Language Model
The 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) 2025
Yunhe Pang, Bo Chen, Fanjin Zhang, Yanghui Rao, Evgeny Kharlamov, Jie Tang
[pdf] [code & data & model]
SAM Decoding: Speculative Decoding via Suffix Automaton
The 63rd Annual Meeting of the Association for Computational Linguistics (ACL) 2025
Yuxuan Hu, Ke Wang, Xiaokang Zhang, Fanjin Zhang, Cuiping Li, Hong Chen, Jing Zhang
[pdf] [code]
OAG-Bench: A Human-Curated Benchmark for Academic Graph Mining
The 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) 2024
Fanjin Zhang, Shijie Shi, Yifan Zhu, Bo Chen, Yukuo Cen, Jifan Yu, Yelin Chen, Lulu Wang, Qingfei Zhao, Yuqing Cheng, Tianyi Han, Yuwei An, Dan Zhang, Weng Lam Tam, Kun Cao, Yunhe Pang, Xinyu Guan, Huihui Yuan, Jian Song, Xiaoyan Li, Yuxiao Dong, and Jie Tang
[pdf] [code] [data] [data challenge]
PST-Bench: Tracing and Benchmarking the Source of Publications
arXiv:2402.16009
Fanjin Zhang, Kun Cao, Yukuo Cen, Jifan Yu, Da Yin, Jie Tang
[pdf] [code & data] [demo]
BOND: Bootstrapping From-Scratch Name Disambiguation with Multi-task Promoting.
The Web Conference (WWW) 2024
Yuqing Cheng, Bo Chen, Fanjin Zhang, and Jie Tang
[pdf] [code]
Web-Scale Academic Name Disambiguation: the WhoIsWho Benchmark, Leaderboard, and Toolkit
The 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) 2023
Bo Chen, Jing Zhang, Fanjin Zhang, Tianyi Han, Yuqing Cheng, XiaoYan Li, Yuxiao Dong, and Jie Tang
[pdf] [code & data]
OAG: Linking Entities across Large-Scale Heterogeneous Knowledge Graphs
IEEE Transactions on Knowledge and Data Engineering (TKDE) 2022
Fanjin Zhang, Xiao Liu, Jie Tang, Yuxiao Dong, Peiran Yao, Jie Zhang, Xiaotao Gu, Yan Wang, Evgeny Kharlamov, Bin Shao, Rui Li, and Kuansan Wang
[pdf] [data]
Understanding WeChat User Preferences and “Wow” Diffusion
IEEE Transactions on Knowledge and Data Engineering (TKDE) 2021
Fanjin Zhang, Jie Tang, Xueyi Liu, Zhenyu Hou, Yuxiao Dong, Jing Zhang, Xiao Liu, Ruobing Xie, Kai Zhuang, Xu Zhang, Leyu Lin, and Philip S. Yu.
[pdf] [code]
Self-supervised Learning: Generative or Contrastive
IEEE Transactions on Knowledge and Data Engineering (TKDE) 2021
Xiao Liu, Fanjin Zhang, Zhenyu Hou, Li Mian, Zhaoyu Wang, Jing Zhang, and Jie Tang
[pdf]
OAG_know: Self-supervised Learning for Linking Knowledge Graphs
IEEE Transactions on Knowledge and Data Engineering (TKDE) 2021
Xiao Liu, Li Mian, Yuxiao Dong, Fanjin Zhang, Jing Zhang, Jie Tang, Peng Zhang, Jibing Gong, and Kuansan Wang
[pdf] [code]
OAG: Toward Linking Large-scale Heterogeneous Entity Graphs
The 25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD Oral) 2019
Fanjin Zhang, Xiao Liu, Jie Tang, Yuxiao Dong, Peiran Yao, Jie Zhang, Xiaotao Gu, Yan Wang, Bin Shao, Rui Li, and Kuansan Wang
[pdf] [code] [data & website]
Name Disambiguation in AMiner: Clustering, Maintenance, and Human in the Loop
The 24th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD Oral) 2018
Yutao Zhang, Fanjin Zhang, Peiran Yao, and Jie Tang
[pdf] [code]
Profiling Web Users Using Big Data
Social Network Analysis and Mining (SNAM) 2018
Xiaotao Gu, Hong Yang, Jie Tang, Jing Zhang, Fanjin Zhang, Debing Liu, Wendy Hall, and Xiao Fu
[pdf]
Experience
2024.01 - Now, Postdoctoral Researcher, Tsinghua University, China
2023.07 - 2023.12, Algorithm Engineer, Zhipu AI, Beijing, China
2017.09 - 2023.06, Ph.D. in Computer Science and Technology, Tsinghua University, China
-
2019.07 - 2020.11, Research Intern, Tencent, Beijing, China
-
2017.05 - 2017.09, Research Intern, Microsoft Research Asia, Beijing, China
2013.09 - 2017.06, B.E. in Computer Science and Technology, Nanjing University, China
Service
PC Member:
2025: KDD
2024: WSDM, AAAI, WWW, KDD
2023: WSDM, AAAI
2020: ECML-PKDD
Reviewer: IEEE Transactions on Knowledge and Data Engineering (TKDE), IEEE Transactions on Neural Networks and Learning Systems (TNNLS), IEEE Transactions on Big Data (TBD), Social Network Analysis and Mining (SNAM), AI Open
Awards
Excellent Graduate Student from Department of Computer Science and Technology, Tsinghua University (2023)
Alumnus Scholarship of Tsinghua University (2021 & 2019)
Second Prize of the National Science and Technology Progress Award (2020)
CAAI Annual Outstanding Scientific and Technological Achievement Award (2020)
National Scholarship (2014)