Explain like I am BM25: Interpreting a Dense Model's Ranked-List with a Sparse Approximation
This work addresses interpretability issues in neural retrieval models for information retrieval practitioners, but it is incremental as it builds on existing sparse retrieval systems.
The paper tackled the problem of poor interpretability in neural retrieval models (NRMs) by introducing equivalent queries as local per-query explanations, comparing this approach with existing methods like RM3-based query expansion in terms of retrieval effectiveness and generated terms.
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.