Improving Retrieval in Theme-specific Applications using a Corpus Topical Taxonomy

SeongKu Kang, Shivam Agarwal, Bowen Jin, Dongha Lee, Hwanjo Yu, Jiawei Han

Keywords: Document retrieval, Topical taxonomy, Theme-specific application

Abstract:

Document retrieval has greatly benefited from the advancements of large-scale pre-trained language models (PLMs) owing to their superior understanding of contextual semantics.
However, their effectiveness is often limited in theme-specific applications for specialized areas or industries, due to unique terminologies, incomplete contexts of user queries, and specialized search intents.
To capture the theme-specific information and improve retrieval, we propose to use a corpus topical taxonomy, which outlines the latent topic structure of the corpus, reflecting user-interested aspects.
We introduce TTER (Topical Taxonomy Enhanced Retrieval) framework, which identifies the central topics of queries and documents with the guidance of the taxonomy, and exploits their topical relatedness to supplement missing contexts.
As a plug-and-play framework, TTER can be flexibly employed to enhance various PLM-based retrievers.
Through extensive quantitative, ablative, and exploratory experiments on two real-world datasets, we ascertain the benefits of using topical taxonomy for retrieval in theme-specific applications and demonstrate the effectiveness of TTER framework.

Track: Search

Submission Guidelines Scope: Yes

Submission Guidelines Blind: Yes

Submission Guidelines Format: Yes

Submission Guidelines Limit: Yes

Submission Guidelines Authorship: Yes

Student Author: No

Submission Number: 1203