OpenIIR: An Open Simulation Platform for Information Retrieval Research

2026-05-10

AI Summary

This work addresses the lack of multi-agent, reproducible, and parameterized simulation platforms in information retrieval research by introducing an open simulation framework powered by large language model–driven virtual agents. The framework enables configurable and reproducible information retrieval experiments across four standardized multi-agent scenarios. Built upon a shared core architecture with plug-in scene interfaces, it integrates a structured world model, retrieval primitives, argument graph generation, and agent ontology modeling, and offers six modular extensions. The system supports systematic comparison of experimental outcomes under varying parameter settings and outputs structured data for direct evaluation, thereby providing a flexible and standardized research infrastructure for investigating open-ended information retrieval problems.

Abstract

OpenIIR runs hundreds of LLM-driven personas as parameterised, reproducible IR research experiments. Researchers configure agents across four kinds of multi-agent study (deliberative panels, social platforms, curated recommender feeds, and evolutionary co-evolution between content producers and credibility detectors) under many priors, rounds, and constraints. Persona budgets, retrieval policies, ranker choices, intervention timings, and mutation rates are declared up front, and the same study can be re-run under different settings to compare outcomes side by side. Every run produces structured outputs (argument graphs, exposure logs, fitness traces, transcripts) that a downstream evaluator can consume directly, and a new study is a 200--400 line plug-in over a shared core (agent runtime, world-model store, retrieval primitives, claim extractor, persona ontology). The contributions are: (i) the shared core; (ii) a type interface for pluggable scenarios; (iii) four released types with reference runs (Panel, Social-Media, Curated-Feed, Multi-Generational); and (iv) six modular extensions sketched against open IR research questions.

Problem

• Information Retrieval
• Multi-agent Simulation
• Reproducible Research
• LLM-driven Personas
• Open Research Platform

Innovation

• multi-agent simulation
• reproducible IR experiments
• pluggable scenarios
• structured evaluation outputs
• LLM-driven personas