Novel application of deep learning to evaluate conversations from a mental health text support service

The Shout text support service supports individuals experiencing mental health distress through anonymous text conversations. As one of the first research projects on the Shout dataset and one of the first significant attempts to apply advanced deep learning to a text messaging service, this project...

Full description

Saved in:
Bibliographic Details
Main Authors: Daniel Cahn, Sarah Yeoh, Lakshya Soni, Ariele Noble, Mark A. Ungless, Emma Lawrance, Ovidiu Şerban
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Natural Language Processing Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2949719124000670
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846123227059847168
author Daniel Cahn
Sarah Yeoh
Lakshya Soni
Ariele Noble
Mark A. Ungless
Emma Lawrance
Ovidiu Şerban
author_facet Daniel Cahn
Sarah Yeoh
Lakshya Soni
Ariele Noble
Mark A. Ungless
Emma Lawrance
Ovidiu Şerban
author_sort Daniel Cahn
collection DOAJ
description The Shout text support service supports individuals experiencing mental health distress through anonymous text conversations. As one of the first research projects on the Shout dataset and one of the first significant attempts to apply advanced deep learning to a text messaging service, this project is a proof-of-concept demonstrating the potential of using deep learning to text messages. Several areas of interest to Shout are identifying texter characteristics, emphasising high suicide-risk participants, and understanding what can make conversations helpful to texters. Therefore, from a mental health perspective, we look at (1) characterising texter demographics strictly based on the vocabulary used throughout the conversation, (2) predicting an individual’s risk of suicide or self-harm, and (3) assessing conversation success by developing robust outcome metrics. To fulfil these aims, a series of Machine Learning models were trained using data from post-conversation surveys to predict the different levels of suicide risk, whether a conversation was helpful, and texter characteristics, such as demographic information. The results show that language models based on Deep Learning significantly improve understanding of this highly subjective dataset. We compare traditional methods and basic meta-features with the latest developments in Transformer-based architectures and showcase the advantages of mental health research.
format Article
id doaj-art-e6f5fe5048ea496d9071a2ca23e974d0
institution Kabale University
issn 2949-7191
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Natural Language Processing Journal
spelling doaj-art-e6f5fe5048ea496d9071a2ca23e974d02024-12-14T06:34:35ZengElsevierNatural Language Processing Journal2949-71912024-12-019100119Novel application of deep learning to evaluate conversations from a mental health text support serviceDaniel Cahn0Sarah Yeoh1Lakshya Soni2Ariele Noble3Mark A. Ungless4Emma Lawrance5Ovidiu Şerban6Department of Computing, Imperial College London, South Kensington Campus, London SW7 2AZ, UKDepartment of Computing, Imperial College London, South Kensington Campus, London SW7 2AZ, UKDepartment of Surgery & Cancer, Imperial College London, St. Mary’s Campus, London W2 1NY, UK; Institute of Global Health Innovation, Imperial College London, South Kensington Campus, London SW7 2AZ, UKMental Health Innovations, London, UKMental Health Innovations, London, UKInstitute of Global Health Innovation, Imperial College London, South Kensington Campus, London SW7 2AZ, UK; Mental Health Innovations, London, UKData Science Institute, Imperial College London, South Kensington Campus, London SW7 2AZ, UK; Department of Computing, Imperial College London, South Kensington Campus, London SW7 2AZ, UK; Corresponding author at: Data Science Institute, Imperial College London, South Kensington Campus, London SW7 2AZ, UK.The Shout text support service supports individuals experiencing mental health distress through anonymous text conversations. As one of the first research projects on the Shout dataset and one of the first significant attempts to apply advanced deep learning to a text messaging service, this project is a proof-of-concept demonstrating the potential of using deep learning to text messages. Several areas of interest to Shout are identifying texter characteristics, emphasising high suicide-risk participants, and understanding what can make conversations helpful to texters. Therefore, from a mental health perspective, we look at (1) characterising texter demographics strictly based on the vocabulary used throughout the conversation, (2) predicting an individual’s risk of suicide or self-harm, and (3) assessing conversation success by developing robust outcome metrics. To fulfil these aims, a series of Machine Learning models were trained using data from post-conversation surveys to predict the different levels of suicide risk, whether a conversation was helpful, and texter characteristics, such as demographic information. The results show that language models based on Deep Learning significantly improve understanding of this highly subjective dataset. We compare traditional methods and basic meta-features with the latest developments in Transformer-based architectures and showcase the advantages of mental health research.http://www.sciencedirect.com/science/article/pii/S2949719124000670Applied mental health researchUnderstanding user conversationsTransformer-based architectures
spellingShingle Daniel Cahn
Sarah Yeoh
Lakshya Soni
Ariele Noble
Mark A. Ungless
Emma Lawrance
Ovidiu Şerban
Novel application of deep learning to evaluate conversations from a mental health text support service
Natural Language Processing Journal
Applied mental health research
Understanding user conversations
Transformer-based architectures
title Novel application of deep learning to evaluate conversations from a mental health text support service
title_full Novel application of deep learning to evaluate conversations from a mental health text support service
title_fullStr Novel application of deep learning to evaluate conversations from a mental health text support service
title_full_unstemmed Novel application of deep learning to evaluate conversations from a mental health text support service
title_short Novel application of deep learning to evaluate conversations from a mental health text support service
title_sort novel application of deep learning to evaluate conversations from a mental health text support service
topic Applied mental health research
Understanding user conversations
Transformer-based architectures
url http://www.sciencedirect.com/science/article/pii/S2949719124000670
work_keys_str_mv AT danielcahn novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT sarahyeoh novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT lakshyasoni novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT arielenoble novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT markaungless novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT emmalawrance novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice
AT ovidiuserban novelapplicationofdeeplearningtoevaluateconversationsfromamentalhealthtextsupportservice