CAPSTONE: Curriculum Sampling for Dense Retrieval with Document Expansion

TL;DR

This paper introduces a curriculum sampling strategy for dense retrieval that improves document representations by progressively aligning generated queries with real queries, leading to better retrieval performance.

Contribution

It proposes a novel curriculum sampling method that enhances query-informed document representations during training, addressing the inconsistency issue in previous approaches.

Findings

Outperforms previous dense retrieval models on multiple datasets

Improves relevance between generated and real queries during training

Enhances the quality of query-informed document representations

Abstract

The dual-encoder has become the de facto architecture for dense retrieval. Typically, it computes the latent representations of the query and document independently, thus failing to fully capture the interactions between the query and document. To alleviate this, recent research has focused on obtaining query-informed document representations. During training, it expands the document with a real query, but during inference, it replaces the real query with a generated one. This inconsistency between training and inference causes the dense retrieval model to prioritize query information while disregarding the document when computing the document representation. Consequently, it performs even worse than the vanilla dense retrieval model because its performance heavily relies on the relevance between the generated queries and the real query.In this paper, we propose a curriculum sampling strategy that utilizes pseudo queries during training and progressively enhances the relevance between the generated query and the real query. By doing so, the retrieval model learns to extend its attention from the document alone to both the document and query, resulting in high-quality query-informed document representations. Experimental results on both in-domain and out-of-domain datasets demonstrate that our approach outperforms previous dense retrieval models.