Rational Retrieval Acts: Leveraging Pragmatic Reasoning to Improve Sparse Retrieval

Arthur Satouf, Gabriel Ben Zenou, Benjamin Piwowarski, Habiboulaye Amadou Boubacar, Pablo Piantanida
5/6/2025
cs.IRcs.CL

Official code: arthur-75/Rational-Retrieval-Acts (7 stars, MIT, last commit 2025-07-16).

Abstract

Current sparse neural information retrieval (IR) methods, and to a lesser extent more traditional models such as BM25, do not take into account the document collection and the complex interplay between different term weights when representing a single document. In this paper, we show how the Rational Speech Acts (RSA), a linguistics framework used to minimize the number of features to be communicated when identifying an object in a set, can be adapted to the IR case -- and in particular to the high number of potential features (here, tokens). RSA dynamically modulates token-document interactions by considering the influence of other documents in the dataset, better contrasting document representations. Experiments show that incorporating RSA consistently improves multiple sparse retrieval models and achieves state-of-the-art performance on out-of-domain datasets from the BEIR benchmark. https://github.com/arthur-75/Rational-Retrieval-Acts

View on arXivView PDF

Code Implementations(1)

arthur-75/Rational-Retrieval-ActsOfficial100%
MIT

Rational Retrieval Acts SIGIR 2025

70Python, Jupyter NotebookApr 21, 20251 years agoMIT

Similar Papers

The Power of Noise: Redefining Retrieval for RAG Systems

80% match1 repo702024

Large Language Model Can Be a Foundation for Hidden Rationale-Based Retrieval

80% match2024

Exploring new Approaches for Information Retrieval through Natural Language Processing

80% match2025

Pseudo Relevance Feedback is Enough to Close the Gap Between Small and Large Dense Retrieval Models

79% match2025

Blending Learning to Rank and Dense Representations for Efficient and Effective Cascades

79% match2025

Dense X Retrieval: What Retrieval Granularity Should We Use?

79% match1 repo1712023

Cite this paper

@article{satouf2025rational,
  title  = {Rational Retrieval Acts: Leveraging Pragmatic Reasoning to Improve Sparse Retrieval},
  author = {Arthur Satouf and Gabriel Ben Zenou and Benjamin Piwowarski and Habiboulaye Amadou Boubacar and Pablo Piantanida},
  year   = {2025},
  eprint = {2505.03676},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2505.03676v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2505.03676.svg)](https://codeofpaper.com/paper/2505.03676)

Discussion