This repository serves as a comprehensive collection of academic papers focused on the application of Large Language Models (LLMs) within the field of Information Retrieval (IR). It is curated to support the survey paper "Large Language Models for Information Retrieval: A Survey," providing researchers and practitioners with a structured overview of LLM-driven techniques across various IR tasks.

How It Works

The repository organizes papers into key IR sub-tasks where LLMs are applied: Query Rewriting, Retrieval (data generation and model enhancement), Re-ranking, Reader models (passive, active, compressors), and Search Agents (static and dynamic). This categorization allows users to quickly identify relevant research for specific LLM-IR integration strategies.

Quick Start & Requirements

• No installation or execution commands are provided as this is a curated list of papers, not executable code.
• Requirements: Access to academic databases or arXiv for paper retrieval.
• Links:
  • Survey Paper: https://arxiv.org/abs/2308.07107
  • BibTex Citation: Provided in the README.

Highlighted Details

• Extensive coverage of LLM applications in IR, including query rewriting, retrieval augmentation, re-ranking, and search agents.
• Regular updates (Version 3 as of Sep 2024) incorporating the latest research trends and methodologies.
• Detailed categorization of papers within each LLM-IR sub-task, facilitating targeted research.
• Includes links to papers and relevant resources like tutorials and other surveys.

Maintenance & Community

• Maintained by RUC-NLPIR.
• Contact emails provided for feedback and advice: yutaozhu94@gmail.com and dou@ruc.edu.cn.
• The project is based on a published survey paper, indicating academic backing.

Licensing & Compatibility

• The repository itself does not appear to have a specific license as it is a collection of links and metadata.
• Individual papers are subject to their respective publication licenses and copyright.

Limitations & Caveats

This repository is a curated list of research papers and does not provide code, models, or benchmarks for direct implementation. Users must independently access and evaluate the cited papers.