Automated Index Selection
Learning-based methods for estimating index benefit and choosing practical database indexes under changing workloads.
Ph.D. Candidate | Fudan University
Chenning Wu
I study AI4DB, automated index selection, and adaptive indexing for modern scalar-vector database workloads.
About
Database systems research for robust, adaptive indexing.
I am a Ph.D. candidate in Big Data and Data Science at the School of Computer Science and Technology, Fudan University. I am currently a visiting researcher in the Data Intelligence Lab at the University of Waterloo, working with Prof. Renée J. Miller. My current research focuses on automated index selection techniques for hybrid scalar-vector scenarios, including uncertainty-aware tuning, learned cost models, filtered ANN search, and robust benefit estimation.
Research Focus
AI4DB, vector search, and index tuning.
Hybrid Scalar-Vector Retrieval
Adaptive techniques for approximate nearest neighbor search with scalar predicates and query-time tuning.
Uncertainty-Aware Optimization
Online tuning strategies that reason about uncertainty at the operator level for dynamic database environments.
Selected Work
Publications and projects
RIB: Robust Learning-based Index Benefit Estimation
Uses bidirectional GNNs and quantile regression to improve robustness in learned index benefit estimation.
UTune: Towards Uncertainty-Aware Online Index Tuning
Proposes an uncertainty-aware operator-level learned model for dynamic online index tuning.
GeoPhrase Tree
Integrates suffix trees with R-trees for efficient frequent phrase queries over spatio-temporal ranges.
Learned Cost Model for Index Tuning on ANN Search
Explores learned vector search models that adapt ef_search during query processing.
Adaptive Indexing for Predicate-Agnostic Filtered ANN Search
Studies adaptive indexing strategies for filtered approximate nearest neighbor workloads.
Education
Academic background
May 2026 - Sep 2026
University of Waterloo (UWaterloo)
Visiting researcher in the Data Intelligence Lab.
2023 - Current
Fudan University
Ph.D., Big Data and Data Science, School of Computer Science and Technology.
2021 - 2023
Fudan University
Master's student, Big Data and Data Science, School of Computer Science and Technology.
2017 - 2021
Fudan University
B.Sc. in Software Engineering, School of Software. GPA 3.68/4.0, Top 5%, Excellent Graduate Award.
Exchange Quarter
University of California, Davis
Exchange program during undergraduate study.
Experience
Industry-facing database research
University-Industry Collaboration with Xiaohongshu
Led a research group in the first Fudan-Xiaohongshu collaboration project, focusing on intelligent database index recommendation research and patent application.
2nd PolarDB Vector Database Optimization Competition
Optimized HNSW index efficiency and improved vector retrieval performance on PolarDB.
Contact