EUVIMEDScopri EUVIMED
Open access

The Most Effective Interventions for Classification Model Development to Predict Chat Outcomes Based on the Conversation Content in Online Suicide Prevention Chats: Machine Learning Approach

Salim Salmi iD, Saskia Mérelle iD, Renske Gilissen iD, Rob van der Mei iD, Sandjai Bhulai iD

JMIR Mental Health · 2024

DOI: 10.2196/57362 ↗

At a glance

Journal article in JMIR Mental Health, 2024. Written by Salim Salmi and 4 co-authors. Institution: Research Department, 113 Suicide Prevention, Amsterdam, Netherlands. Major MEDLINE topics: Machine Learning, Suicide Prevention, Hotlines. Cited 2 times so far (Crossref); 34 references. The full text is freely available in PubMed Central.

AbstractSubjectsDati bibliograficiCiteRelated studies

Abstract (original language)

Background

For the provision of optimal care in a suicide prevention helpline, it is important to know what contributes to positive or negative effects on help seekers. Helplines can often be contacted through text-based chat services, which produce large amounts of text data for use in large-scale analysis.

Objective

We trained a machine learning classification model to predict chat outcomes based on the content of the chat conversations in suicide helplines and identified the counsellor utterances that had the most impact on its outputs.

Methods

From August 2021 until January 2023, help seekers (N=6903) scored themselves on factors known to be associated with suicidality (eg, hopelessness, feeling entrapped, will to live) before and after a chat conversation with the suicide prevention helpline in the Netherlands (113 Suicide Prevention). Machine learning text analysis was used to predict help seeker scores on these factors. Using 2 approaches for interpreting machine learning models, we identified text messages from helpers in a chat that contributed the most to the prediction of the model.

Results

According to the machine learning model, helpers’ positive affirmations and expressing involvement contributed to improved scores of the help seekers. Use of macros and ending the chat prematurely due to the help seeker being in an unsafe situation had negative effects on help seekers.

Conclusions

This study reveals insights for improving helpline chats, emphasizing the value of an evocative style with questions, positive affirmations, and practical advice. It also underscores the potential of machine learning in helpline chat analysis.

Subjects

MeSH terms (MEDLINE)

Machine LearningSuicide PreventionHotlinesHumansMaleFemaleAdultNetherlandsMiddle AgedText Messaging

Author keywords

BERTLLMartificial intelligencebidirectional encoder representations from transformersclassificationconversationsexplainable AIhelplineinterpretable AIlarge language modelsmachine learningnatural language processingsuicidalitysuicidesuicide helpline

Authors and institutions

  • Research Department, 113 Suicide Prevention, Amsterdam, Netherlands.
  • Department of Stochastics, Centrum Wiskunde & Informatica, Amsterdam, Netherlands.
  • Department of Mathematics, Vrije Universiteit Amsterdam, Amsterdam, Netherlands.

ORCID profiles: Salim Salmi, Saskia Mérelle, Renske Gilissen, Rob van der Mei, Sandjai Bhulai

Cite

APA

Salmi, S., Mérelle, S., Gilissen, R., van der Mei, R., & Bhulai, S. (2024). The Most Effective Interventions for Classification Model Development to Predict Chat Outcomes Based on the Conversation Content in Online Suicide Prevention Chats: Machine Learning Approach. JMIR Mental Health. https://doi.org/10.2196/57362

Vancouver

Salmi S, Mérelle S, Gilissen R, van der Mei R, Bhulai S. The Most Effective Interventions for Classification Model Development to Predict Chat Outcomes Based on the Conversation Content in Online Suicide Prevention Chats: Machine Learning Approach. JMIR Mental Health. 2024. doi:10.2196/57362

BibTeX
@article{Salmi2024,
  author = {Salmi, Salim and Mérelle, Saskia and Gilissen, Renske and van der Mei, Rob and Bhulai, Sandjai},
  title = {The Most Effective Interventions for Classification Model Development to Predict Chat Outcomes Based on the Conversation Content in Online Suicide Prevention Chats: Machine Learning Approach},
  journal = {JMIR Mental Health},
  year = {2024},
  doi = {10.2196/57362}
}
Start for free →