Zeyu Wang (王泽宇)

Zeyu Wang (王泽宇)

Fudan University & Université Paris Cité Address: 705G, 45 Rue des Saints-Pères, 75006 Paris

I am now a forth-year PhD candidate at School of Computer Science, Fudan University. My Ph.D. thesis is jointly-supervised by Prof. Peng Wang at Fudan Univ. and Prof. Themis Palpanas at Université Paris Cité. I enjoy working with labmates of DSM Group and diNo Research group at the same time. Before that, I got my bachelor’s degree in computer science from Fudan University in 2021.

I have published papers in top conferences and journals of database and data mining, including SIGMOD, VLDB, VLDB Journal and TKDE. I also serve as the reviewer of peer-reviewed journals including TKDE.

My research interest lies in the management and analyses of massive complex data, including:

1) High-dimensional vector indexing for approximate nearest neighnor search (ANNS)

2) Data series / time series similarity search and analyses

3) Other important issues on high-dimensional data management (e.g., storage, compression)

4) DB + AI: RAG, learned index, learned query optimizer, learned index tuning, etc.

Welcome to contact me via [zeyuwang21@m.fudan.edu.cn] if you are interested in me and my research works.

Welcome to subscribe my Medium to discuss the exciting ideas in data management!

My Chinese tech blog records my detailed comments on academic papers and advanced techniques.

My MBTI is INTJ.

News

  • 2024.8 😄 Very glad to celebrate the 50th anniversary of VLDB in Guangzhou, China with friends in database area!
  • 2024.8 🎉🎉🎉 One new work gets accepted by PVLDB, which studies the cost estimation, hardness measure and the stress-test workload generation of querying graph-based ANN indexes.
  • 2024.8 🎉🎉🎉 Our new work DumpyOS, a state-of-the-art parallel time series index on NVMe SSD, gets accepted in VLDB Journal!
  • 2024.4 🎉🎉🎉 Our new work CIVET, a state-of-the-art time series subsequence matching index, gets accepted in VLDB 2024!

Publications

  • Zeyu Wang, Qitong Wang, Xiaoxing Cheng, Peng Wang, Themis Palpanas, and Wei Wang.

    $Steiner$-Hardness: A Query Hardness Measure for Graph-Based ANN Indexes.

    PVLDB, 17(13) 2024, will be presented in London, UK, 2025 (PDF, code)

  • Zeyu Wang, Qitong Wang, Peng Wang, Themis Palpanas, and Wei Wang.

    DumpyOS: A Data-Adaptive Multi-ary Index for Scalable Data Series Similarity Search.

    The VLDB Journal, Aug. 2024 (PDF, code)

  • Haoran Xiong, Hang Zhang, Zeyu Wang, Zhenying He, Peng Wang, and X. Sean Wang.

    CIVET: Exploring Compact Index for Variable-Length Subsequence Matching on Time Series.

    PVLDB, 17(9): 2123-2135, Aug. 2024, Guangzhou, China. (PDF, code)

  • Zeyu Wang, Haoran Xiong, Zhenying He, Peng Wang, and Wei Wang.

    Dimensionality-Reduction Techniques for Approximate Nearest Neighbor Search: A Survey and Evaluation.

    IEEE Data Engineering Bulletin (invited paper) 48(3), Sep. 2024. (PDF, code, to appear)

  • Zeyu Wang, Zhenying He, Peng Wang, Yang Wang, and Wei Wang.

    Static and Streaming Discovery of Maximal Linear Representation Between Time Series.

    IEEE Transactions on Knowledge and Data Engineering (TKDE), vol. 36, no. 1, pp. 401-415, Jan. 2024. (website, code)

  • Zeyu Wang, Peng Wang, Themis Palpanas, and Wei Wang.

    Graph- and Tree-based Indexes for High-dimensional Vector Similarity Search: Analyses, Comparisons, and Future Directions.

    IEEE Data Engineering Bulletin (invited paper) 47(3), Sep. 2023. (PDF)

  • Zeyu Wang, Qitong Wang, Peng Wang, Themis Palpanas, and Wei Wang.

    Dumpy: A Compact and Adaptive Index for Large Data Series Collections.

    Proceedings of the ACM Management of Data (PACMMOD) Journal 1(1), presented at ACM SIG International conference on Management of Data / Principles of Database Systems (SIGMOD/PODS), June 2023, Seattle, WA, USA. (PDF, slides, video, code)

  • Hanbo Zhang, Peng Wang, Zicheng Fang, Zeyu Wang, and Wei Wang.

    ELIS++: a shapelet learning approach for accurate and efficient time series classification.

    World Wide Web (WWWJ) 24, 511–539 March 2021. (website)

Activities

Talks

  • Efficient and Reliable Automatic High-dimensional Vector Indexes. In Ant Group, July 2024.
  • A Revisit and the New Progress of Graph-based High-dimensional Vector Search. In DataWhale, online, May 2023. video (in Mandarin)
  • Similarity Search: From Time Series to High-Dimensional Vectors. In Zilliz, Aug 2022.

Teaching Assistants

  • Big Data Mining, Spring 2022, with Prof. Peng Wang
  • Database System Implementation, Fall 2022, with Prof. Peng Wang and Prof. Wei Wang

Interns

  • Research and Development at Baidu (2024.7 - 2024.11)
    • Working on the content relation team of Chinese largest search engine Baidu, and optimizing the similarity search engine Puck
  • Researcher at Ant Group (2024.7)
  • Research Engineer at Zilliz (2022.8 - 2023.1)
    • Working at the research group led by Dr. Xiaomeng Yi for learned graph-based ANN indexes and graph indexes in streaming
  • BigData Engineer at Construct Tech (2020.1 - 2021.1)
    • Leading the data team to build the big data platform to empower intelligent analyses, recommendation, and social matches.

Awards

  • Top-50 team on Oceanbase Database Competition in 2022. (also the National College Student Computer System ability Competition in China, from 2023)
  • Outstanding Graduate of Fudan University in 2021
  • Gold Award Team at iGEM (intertional Genetically Engineered Machine Competition) in 2019, Boston, USA. Presented the modeling of how the biological therapy take effect for lactose intolerance in the Giant Jamboree.
  • National Second Prize of Undergraduate Mathematical Contest in Modeling Competition in China, in 2019