Explain like I am BM25: Interpreting a Dense Model's Ranked-List with a Sparse Approximation

2023/04/25 by Michael Llordes, Debasis Ganguly, Llordes, Michael +5 · 4 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · #Biomedical Text Mining and Ontologies #Computation and Language (cs.CL) #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2304.12631

openalex publication_date 2023/04/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Neural retrieval models (NRMs) have been shown to outperform their statistical counterparts owing to their ability to capture semantic meaning via dense document representations. These models, however, suffer from poor interpretability as they do not rely on explicit term matching. As a form of local per-query explanations, we introduce the notion of equivalent queries that are generated by maximizing the similarity between the NRM's results and the result set of a sparse retrieval system with the equivalent query. We then compare this approach with existing methods such as RM3-based query expansion and contrast differences in retrieval effectiveness and in the terms generated by each approach.

Cited by

• A Survey on Retrieval And Structuring Augmented Generation with Large Language Models
• Explainable Information Retrieval in the Audit Domain
• Decoding Dense Embeddings: Sparse Autoencoders for Interpreting and Discretizing Dense Retrieval
• IterKey: Iterative Keyword Generation with LLMs for Enhanced Retrieval Augmented Generation

Related

• Evaluating the Explainability of Neural Rankers
• Interpreting search result rankings through intent modeling
• Improving Document Retrieval Coherence for Semantically Equivalent Queries
• Discovering Biases in Information Retrieval Models Using Relevance Thesaurus as Global Explanation
• Faster Learned Sparse Retrieval with Guided Traversal