DREditor: An Time-efficient Approach for Building a Domain-specific Dense Retrieval Model

TL;DR

DREditor is a novel, time-efficient method for customizing dense retrieval models to specific domains by calibrating output embeddings through a linear mapping, significantly reducing calibration time while maintaining or improving performance.

Contribution

The paper introduces DREditor, a new embedding calibration approach that enables rapid domain adaptation of dense retrieval models without extensive fine-tuning.

Findings

DREditor achieves 100-300x time efficiency improvement.

It maintains or surpasses the retrieval performance of traditional fine-tuning.

The approach is effective across various datasets, models, and hardware.

Abstract

Deploying dense retrieval models efficiently is becoming increasingly important across various industries. This is especially true for enterprise search services, where customizing search engines to meet the time demands of different enterprises in different domains is crucial. Motivated by this, we develop a time-efficient approach called DREditor to edit the matching rule of an off-the-shelf dense retrieval model to suit a specific domain. This is achieved by directly calibrating the output embeddings of the model using an efficient and effective linear mapping. This mapping is powered by an edit operator that is obtained by solving a specially constructed least squares problem. Compared to implicit rule modification via long-time finetuning, our experimental results show that DREditor provides significant advantages on different domain-specific datasets, dataset sources, retrieval models, and computing devices. It consistently enhances time efficiency by 100-300 times while maintaining comparable or even superior retrieval performance. In a broader context, we take the first step to introduce a novel embedding calibration approach for the retrieval task, filling the technical blank in the current field of embedding calibration. This approach also paves the way for building domain-specific dense retrieval models efficiently and inexpensively.