Language Transfer for Early Warning of Epidemics from Social Media

TL;DR

This paper investigates multilingual transfer learning methods for early epidemic detection from social media, focusing on Japanese and exploring translation-based and zero-shot approaches to overcome data scarcity.

Contribution

It evaluates the effectiveness of machine translation and multilingual models for cross-lingual epidemic warning systems, especially for low-resource languages.

Findings

Chinese-Japanese language pair outperforms English-Japanese.

Machine translated data enhances model performance when combined with limited target language data.

Multilingual models enable zero-shot transfer for epidemic detection across languages.

Abstract

Statements on social media can be analysed to identify individuals who are experiencing red flag medical symptoms, allowing early detection of the spread of disease such as influenza. Since disease does not respect cultural borders and may spread between populations speaking different languages, we would like to build multilingual models. However, the data required to train models for every language may be difficult, expensive and time-consuming to obtain, particularly for low-resource languages. Taking Japanese as our target language, we explore methods by which data in one language might be used to build models for a different language. We evaluate strategies of training on machine translated data and of zero-shot transfer through the use of multilingual models. We find that the choice of source language impacts the performance, with Chinese-Japanese being a better language pair than English-Japanese. Training on machine translated data shows promise, especially when used in conjunction with a small amount of target language data.