Wolverine: Highly Efficient Monotonic Search Path Repair for Graph-based ANN Index Updates
Summary: Wolverine: a monotonic search-path repair framework for dynamic graph-based ANN indices that fixes broken monotonic paths by adding in-edges to out-neighbors of deleted nodes to preserve connectivity and recall. Wolverine+ (2‑hop restriction) and Wolverine++ (quality-driven candidate selection) speed deletions up to 11× and maintain steadier recall vs. prior dynamic ANN methods across 9 real datasets. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
Authors
- 1. Dawei Liu (Huazhong University of Science and Technology)
- 2. Bolong Zheng (Huazhong University of Science and Technology)
- 3. Ziyang Yue (Huazhong University of Science and Technology)
- 4. Fuhao Ruan (Huazhong University of Science and Technology)
- 5. Xiaofang Zhou (Hong Kong University of Science and Technology)
- 6. Christian S. Jensen (Aalborg University)
BibTeX Citation
@article{liu_vldb25,
title = {{Wolverine: Highly Efficient Monotonic Search Path Repair for Graph-based ANN Index Updates}},
author = {Liu, Dawei and Zheng, Bolong and Yue, Ziyang and Ruan, Fuhao and Zhou, Xiaofang and Jensen, Christian S.},
journal = {PVLDB},
series = {{VLDB} '25},
volume = {18},
number = {7},
pages = {2268--2280},
doi = {10.14778/3734839.3734860},
url = {https://doi.org/10.14778/3734839.3734860},
year = {2025}
}
Incoming Citations (Sorted by Pagerank)
Showing 2 of 2 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,251 | GPS: Revisiting the Data Layout for Disk-based High-Dimensional Vector Search | 2026 | SIGMOD | 5.093636e-05 |
| 10,299 | Through the Lens of Hubness: A Revisit on Graph-Based Approximate Nearest Neighbor Search: [Experiments & Analysis] | 2026 | SIGMOD | 5.093636e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 11 of 11 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
Previous
Page 1 / 1
Next