A Multi-Agent Perspective on Modern Information Retrieval
Haya Nachimovsky, Moshe Tennenholtz, Oren Kurland
TL;DR
The paper argues that modern information retrieval must be rethought in the presence of autonomous agents that generate queries, documents, and rankings. It formalizes a multi-agent IR framework and analyzes how alignment across agent types (query, document, ranker) and competitive incentives shape retrieval effectiveness. Through a suite of lexical, semantic, and LLM-based agents and simulation-based evaluation, the authors show that misalignment degrades performance and that corpus composition (human vs. LLM content) critically affects ranking outcomes. The work highlights the need for simulation-based evaluation, robust alignment mechanisms, and new axioms to model and mitigate corpus effects in multi-agent retrieval scenarios.
Abstract
The rise of large language models (LLMs) has introduced a new era in information retrieval (IR), where queries and documents that were once assumed to be generated exclusively by humans can now also be created by automated agents. These agents can formulate queries, generate documents, and perform ranking. This shift challenges some long-standing IR paradigms and calls for a reassessment of both theoretical frameworks and practical methodologies. We advocate for a multi-agent perspective to better capture the complex interactions between query agents, document agents, and ranker agents. Through empirical exploration of various multi-agent retrieval settings, we reveal the significant impact of these interactions on system performance. Our findings underscore the need to revisit classical IR paradigms and develop new frameworks for more effective modeling and evaluation of modern retrieval systems.