XLTime: A Cross-Lingual Knowledge Transfer Framework for Temporal Expression Extraction

TL;DR

XLTime is a multilingual framework that enhances temporal expression extraction across languages by leveraging cross-lingual knowledge transfer with pre-trained language models, significantly outperforming previous methods.

Contribution

It introduces XLTime, a novel multi-task learning framework that improves temporal expression extraction in multiple languages using cross-lingual transfer from English and within non-English languages.

Findings

Outperforms previous SOTA methods on French, Spanish, Portuguese, and Basque.

Reduces the data scarcity problem in non-English TEE.

Closes the gap with handcrafted HeidelTime method.

Abstract

Temporal Expression Extraction (TEE) is essential for understanding time in natural language. It has applications in Natural Language Processing (NLP) tasks such as question answering, information retrieval, and causal inference. To date, work in this area has mostly focused on English as there is a scarcity of labeled data for other languages. We propose XLTime, a novel framework for multilingual TEE. XLTime works on top of pre-trained language models and leverages multi-task learning to prompt cross-language knowledge transfer both from English and within the non-English languages. XLTime alleviates problems caused by a shortage of data in the target language. We apply XLTime with different language models and show that it outperforms the previous automatic SOTA methods on French, Spanish, Portuguese, and Basque, by large margins. XLTime also closes the gap considerably on the handcrafted HeidelTime method.