Building Interpretable and Reliable Open Information Retriever for New Domains Overnight

TL;DR

This paper introduces an interpretable and reliable open-domain information retrieval system that improves cross-domain performance by using entity/event linking and query decomposition, especially for new domains.

Contribution

The work presents a novel IR pipeline combining entity/event linking and query decomposition to enhance interpretability and effectiveness across domains.

Findings

Significantly improves passage coverage and denotation accuracy.

Outperforms state-of-the-art dense retrieval models on multiple benchmarks.

Enhances interpretability and reliability for IR in new domains.

Abstract

Information retrieval (IR) or knowledge retrieval, is a critical component for many down-stream tasks such as open-domain question answering (QA). It is also very challenging, as it requires succinctness, completeness, and correctness. In recent works, dense retrieval models have achieved state-of-the-art (SOTA) performance on in-domain IR and QA benchmarks by representing queries and knowledge passages with dense vectors and learning the lexical and semantic similarity. However, using single dense vectors and end-to-end supervision are not always optimal because queries may require attention to multiple aspects and event implicit knowledge. In this work, we propose an information retrieval pipeline that uses entity/event linking model and query decomposition model to focus more accurately on different information units of the query. We show that, while being more interpretable and reliable, our proposed pipeline significantly improves passage coverages and denotation accuracies across five IR and QA benchmarks. It will be the go-to system to use for applications that need to perform IR on a new domain without much dedicated effort, because of its superior interpretability and cross-domain performance.