Investigating Keyphrase Indexing with Text Denoising

TL;DR

This study evaluates how pairing a state-of-the-art keyphrase indexer with text denoising affects indexing performance across multiple domains, showing that training on denoised texts can match or outperform full-text training.

Contribution

It introduces the use of text denoising for training keyphrase indexers and demonstrates its effectiveness across diverse scientific domains.

Findings

Denoised-text-trained indexer performs as well as or better than full-text-trained indexer.

Training on denoised texts reduces noise and improves indexing accuracy.

Results are consistent across food, physics, and biomedical corpora.

Abstract

In this paper, we report on indexing performance by a state-of-the-art keyphrase indexer, Maui, when paired with a text extraction procedure called text denoising. Text denoising is a method that extracts the denoised text, comprising the content-rich sentences, from full texts. The performance of the keyphrase indexer is demonstrated on three standard corpora collected from three domains, namely food and agriculture, high energy physics, and biomedical science. Maui is trained using the full texts and denoised texts. The indexer, using its trained models, then extracts keyphrases from test sets comprising full texts, and their denoised and noise parts (i.e., the part of texts that remains after denoising). Experimental findings show that against a gold standard, the denoised-text-trained indexer indexing full texts, performs either better than or as good as its benchmark performance produced by a full-text-trained indexer indexing full texts.