Improving Mental Health using Machine Learning to Assist Humans in the Moderation of Forum Posts

Dong Wang, Julie Weeds, Ian Comley

2020

Abstract

This work investigates the potential for the application of machine learning and natural language processing technology in an online application designed to help teenagers talk about their mental health issues. Specifically, we investigate whether automatic classification methods can be applied with sufficient accuracy to assist humans in the moderation of posts and replies to an online forum. Using real data from an existing application, we outline the specific problems of lack of data, class imbalance and multiple rejection reasons. We investigate a number of machine learning architectures including a state-of-the-art transfer learning architecture, BERT, which has performed well elsewhere despite limited training data, due to its use of pre-training on a very large general corpus. Evaluating on real data, we demonstrate that further large performance gains can be made through the use of automatic data augmentation techniques (synonym replacement, synonym insertion, random swap and random deletion). Using a combination of data augmentation and transfer learning, performance of the automatic classification rivals human performance at the task, thus demonstrating the feasibility of deploying these techniques in a live system.

Paper Citation

in Harvard Style

Wang D., Weeds J. and Comley I. (2020). Improving Mental Health using Machine Learning to Assist Humans in the Moderation of Forum Posts. In Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 5: HEALTHINF; ISBN 978-989-758-398-8, SciTePress, pages 187-197. DOI: 10.5220/0008988401870197

in Bibtex Style

@conference{healthinf20,
author={Dong Wang and Julie Weeds and Ian Comley},
title={Improving Mental Health using Machine Learning to Assist Humans in the Moderation of Forum Posts},
booktitle={Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 5: HEALTHINF},
year={2020},
pages={187-197},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0008988401870197},
isbn={978-989-758-398-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 5: HEALTHINF
TI - Improving Mental Health using Machine Learning to Assist Humans in the Moderation of Forum Posts
SN - 978-989-758-398-8
AU - Wang D.
AU - Weeds J.
AU - Comley I.
PY - 2020
SP - 187
EP - 197
DO - 10.5220/0008988401870197
PB - SciTePress