Keyword Extraction from Short Texts with a Text-To-Text Transfer Transformer

TL;DR

This paper evaluates the plT5 model for keyword extraction from short texts, demonstrating its effectiveness on scientific abstracts and cross-domain texts, and introduces a new Polish scientific metadata corpus.

Contribution

It presents a new Polish scientific metadata corpus and demonstrates the effectiveness of the plT5 model for keyword extraction across multiple domains.

Findings

plT5kw outperforms other methods in keyword extraction

The new POSMAC corpus facilitates evaluation of keyword extraction models

plT5kw shows promise in cross-domain text labelling scenarios

Abstract

The paper explores the relevance of the Text-To-Text Transfer Transformer language model (T5) for Polish (plT5) to the task of intrinsic and extrinsic keyword extraction from short text passages. The evaluation is carried out on the new Polish Open Science Metadata Corpus (POSMAC), which is released with this paper: a collection of 216,214 abstracts of scientific publications compiled in the CURLICAT project. We compare the results obtained by four different methods, i.e. plT5kw, extremeText, TermoPL, KeyBERT and conclude that the plT5kw model yields particularly promising results for both frequent and sparsely represented keywords. Furthermore, a plT5kw keyword generation model trained on the POSMAC also seems to produce highly useful results in cross-domain text labelling scenarios. We discuss the performance of the model on news stories and phone-based dialog transcripts which represent text genres and domains extrinsic to the dataset of scientific abstracts. Finally, we also attempt to characterize the challenges of evaluating a text-to-text model on both intrinsic and extrinsic keyword extraction.