Optimal fusion weights for search vary by query in intervals
Optimal Fusion Weights Form Intervals, Not Points: A Latency-Aware Study of Query-Adaptive WRRF
Information Retrieval
Summary
When combining two ways to find information—one focusing on exact matches and another on broader meanings—deciding how much to trust each method is important. This paper shows that for each search, the best trust level isn't a single perfect number but a range of good choices. The authors tested many ways to pick these trust ranges and found many improved results compared to a fixed mixing approach. They also created a guide to help choose the best method based on factors like hardware and speed needs.
What this means in practice
- •For search system engineers: Design search engines that adaptively mix retrieval methods based on query-specific ranges of weights instead of fixed values to improve result relevance.
- •For information retrieval platform developers: Select fusion strategies that balance retrieval quality and processing speed based on hardware capabilities and latency requirements for hybrid search applications.
Authors
Iris Ma, Zahra Shekarchi, Stuart Mcilroy, Mahtab Farrokh, Fattane Zarrinkalam, Joel Stremmel
Abstract
Reciprocal Rank Fusion (RRF) is widely used to combine sparse and dense retrievers in hybrid information retrieval, but a critical limitation is that it assigns a fixed, uniform weight to each retriever regardless of the query. Query-aware methods treat the per-query optimal weight as a single point, an assumption whose empirical validity has not been examined. We present an empirical study of eight dynamic fusion weight prediction strategies across four datasets and four sparse--dense retriever pairs. When we compare these approaches to static fusion and a per-query oracle that bounds achievable performance, several findings emerge. First, the per-query optimal fusion weight is typically an interval rather than a point. Queries can have multiple optimal intervals, and these can take on a variety of shapes. Second, 74 of 128 dynamic fusion methods tested were significantly better than standard RRF after false discovery rate correction. Third, dynamic fusion methods exhibit differing quality and latency characteristics that should be considered when choosing an appropriate fusion method. To address this last finding, we present a tiered decision framework that maps practitioner constraints, dataset properties, label availability, hardware, and latency budget to an appropriate method class. We release the code for our experiments at https://github.com/thomsonreuters/query-based-rrf.