Curator: Efficient Vector Search with Low-Selectivity Filters

TL;DR

Curator introduces a dual-index system that enhances vector search efficiency under low-selectivity filters by partitioning data and adapting indexes to label distributions, significantly reducing query latency.

Contribution

It proposes a novel partition-based index architecture that complements graph indexes, improving low-selectivity filtered ANNS performance with minimal overhead.

Findings

Reduces low-selectivity query latency by up to 20.9x

Maintains low memory overhead with only 4.3% increase

Supports incremental updates and complex predicates

Abstract

Embedding-based dense retrieval has become the cornerstone of many critical applications, where approximate nearest neighbor search (ANNS) queries are often combined with filters on labels such as dates and price ranges. Graph-based indexes achieve state-of-the-art performance on unfiltered ANNS but encounter connectivity breakdown on low-selectivity filtered queries, where qualifying vectors become sparse and the graph structure among them fragments. Recent research proposes specialized graph indexes that address this issue by expanding graph degree, which incurs prohibitively high construction costs. Given these inherent limitations of graph-based methods, we argue for a dual-index architecture and present Curator, a partition-based index that complements existing graph-based approaches for low-selectivity filtered ANNS. Curator builds specialized indexes for different labels within a shared clustering tree, where each index adapts to the distribution of its qualifying vectors to ensure efficient search while sharing structure to minimize memory overhead. The system also supports incremental updates and handles arbitrary complex predicates beyond single-label filters by efficiently constructing temporary indexes on the fly. Our evaluation demonstrates that integrating Curator with state-of-the-art graph indexes reduces low-selectivity query latency by up to 20.9x compared to pre-filtering fallback, while increasing construction time and memory footprint by only 5.5% and 4.3%, respectively.