Incremental IVF Index Maintenance for Streaming Vector Search

TL;DR

This paper introduces Ada-IVF, an incremental index maintenance method for vector search that improves update throughput and maintains search quality in streaming data scenarios.

Contribution

Ada-IVF provides an adaptive maintenance policy and local re-clustering mechanism for efficient incremental IVF index updates.

Findings

Achieves 2x to 5x higher update throughput than existing methods.

Maintains search quality with less costly index updates.

Effective for streaming and dynamic data workloads.

Abstract

The prevalence of vector similarity search in modern machine learning applications and the continuously changing nature of data processed by these applications necessitate efficient and effective index maintenance techniques for vector search indexes. Designed primarily for static workloads, existing vector search indexes degrade in search quality and performance as the underlying data is updated unless costly index reconstruction is performed. To address this, we introduce Ada-IVF, an incremental indexing methodology for Inverted File (IVF) indexes. Ada-IVF consists of 1) an adaptive maintenance policy that decides which index partitions are problematic for performance and should be repartitioned and 2) a local re-clustering mechanism that determines how to repartition them. Compared with state-of-the-art dynamic IVF index maintenance strategies, Ada-IVF achieves an average of 2x and up to 5x higher update throughput across a range of benchmark workloads.