SERM: Self-Evolving Relevance Model with Agent-Driven Learning from Massive Query Streams

TL;DR

SERM is a novel self-evolving relevance model that uses multi-agent modules to detect informative samples and generate reliable labels, improving search relevance in large-scale query streams.

Contribution

The paper introduces SERM, a self-evolving relevance model with multi-agent modules for sample mining and label annotation, addressing challenges in dynamic query streams.

Findings

SERM achieves significant performance improvements in large-scale industrial settings.

Extensive offline and online evaluations validate the effectiveness of SERM.

SERM effectively detects distributional shifts and identifies informative samples.

Abstract

Due to the dynamically evolving nature of real-world query streams, relevance models struggle to generalize to practical search scenarios. A sophisticated solution is self-evolution techniques. However, in large-scale industrial settings with massive query streams, this technique faces two challenges: (1) informative samples are often sparse and difficult to identify, and (2) pseudo-labels generated by the current model could be unreliable. To address these challenges, in this work, we propose a Self-Evolving Relevance Model approach (SERM), which comprises two complementary multi-agent modules: a multi-agent sample miner, designed to detect distributional shifts and identify informative training samples, and a multi-agent relevance annotator, which provides reliable labels through a two-level agreement framework. We evaluate SERM in a large-scale industrial setting, which serves billions of user requests daily. Experimental results demonstrate that SERM can achieve significant performance gains through iterative self-evolution, as validated by extensive offline multilingual evaluations and online testing.