About Me

I am a final-year PhD candidate in Computer Science at the University of Amsterdam (INDElab), advised by Prof. Dr. Paul Groth and Dr. Klim Zaporojets. My defense is planned for March 2027.

I work on machine learning for knowledge graphs, where the data keeps changing and many entities look alike. I build models that add signals which are usually left out, such as time and how the graph changes, so that entity linking, link prediction and recommendation stay accurate on sparse, long-tail and look-alike cases. The same problem appears in industry as entity resolution under data drift.

Research interests: multi-modal knowledge graphs, entity linking and resolution, link prediction, recommendation, large language models.

News

  • 2026.10 Fina Polat presented our paper Select, Don’t Train at ISWC 2026 in Bari.
  • 2026.08 TimeRoute is on arXiv.
  • 2026.07 Select, Don’t Train was accepted at ISWC 2026.
  • 2026.07 Time Imprint is on arXiv.
  • 2026.07 Presented Fusion Training for Mathematical Generalization in Large Language Models at the ACL 2026 Student Research Workshop in San Diego.
  • 2026.05 Presented Beyond Images at ESWC 2026 in Dubrovnik.
MoreLess
  • 2026.05 Presented Graph-TempCZ at LREC 2026 in Palma de Mallorca, and had many good conversations there about how papers mention software.
  • 2026.04 Fusion Training for Mathematical Generalization in Large Language Models was accepted at the ACL 2026 Student Research Workshop.
  • 2026.02 Beyond Images was accepted at ESWC 2026. Many thanks to my co-authors.
  • 2026.02 Graph-TempCZ was accepted at LREC 2026.
  • 2025.09 Our survey on large language models for data challenges in graphs was accepted by Expert Systems with Applications.
  • 2024.11 Presented our work on entity linking and co-occurrence networks at EKAW 2024, here in Amsterdam.
  • 2024.10 Presented CYCLE at CIKM 2024 in Boise.
  • 2024.10 Presented TIGER at ECAI 2024 in Santiago de Compostela.
  • 2024.09 Our paper on entity linking in co-occurrence networks was accepted at EKAW 2024.
  • 2024.07 CYCLE was accepted at CIKM 2024.
  • 2024.07 TIGER was accepted at ECAI 2024.
  • 2022.09 Joined INDElab at the University of Amsterdam to start my PhD.

Education

  • 2022 - Present PhD in Computer Science (defense planned for March 2027), INDElab, Faculty of Science, University of Amsterdam (UvA), the Netherlands. Supervisors: Prof. Dr. Paul Groth, Dr. Klim Zaporojets.
  • 2019 - 2022 MEng in Control Engineering, Faculty of Information Technology, Beijing University of Technology (BJUT), China. Supervisor: Prof. Dr. Yong Zhang.

Awards & Service

Awards

  • 2026 Conference travel grant for ACL 2026.
  • 2022 Outstanding Master’s Thesis, Beijing University of Technology.

Service

  • Reviewer / sub-reviewer: CIKM, ECAI, ESWC, ECML PKDD.
  • 2020 - 2021 Teaching assistant, Data Engineering (BJUT).

Skills

Programming

Python, PostgreSQL, Bash, Git, Linux

ML tools

PyTorch, Hugging Face, scikit-learn, NumPy, pandas

Methods

Large language models, Vision-language models, Graph neural networks, Contrastive learning, Entity linking and resolution, Recommendation

Languages

Chinese (native), English (fluent, IELTS 7.0), Dutch (elementary, A2)