Compass: General Filtered Search across Vector and Structured Data
Chunxiao Ye, Xiao Yan, Eric Lo

TL;DR
Compass is a unified framework enabling efficient, general filtered search across vector and structured data without specialized indices, supporting complex queries and maintaining high performance.
Contribution
It introduces a cooperative query execution strategy that integrates existing index structures for vector and relational data, enabling flexible, robust hybrid search.
Findings
Outperforms NaviX across diverse hybrid query workloads.
Matches single-attribute index throughput with full generality.
Maintains robustness with highly-selective and multi-attribute filters.
Abstract
The increasing prevalence of hybrid vector and relational data necessitates efficient, general support for queries that combine high-dimensional vector search with complex relational filtering. However, existing filtered search solutions are fundamentally limited by specialized indices, which restrict arbitrary filtering and hinder integration with general-purpose DBMSs. This work introduces \textsc{Compass}, a unified framework that enables general filtered search across vector and structured data without relying on new index designs. Compass leverages established index structures -- such as HNSW and IVF for vector attributes, and B+-trees for relational attributes -- implementing a principled cooperative query execution strategy that coordinates candidate generation and predicate evaluation across modalities. Uniquely, Compass maintains generality by allowing arbitrary conjunctions,…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
