Abstract—Suicide has become one of the leading causes of death worldwide. It is a serious public health problem, and the right prompt response can mitigate it. Therefore, identifying individuals with suicide risk and offering immediate counseling to everyone that might need it is a crucial step. In this research, we utilize personal narratives collected through the popular social media website (Reddit) to build a model suitable for predicting suicide ideation in a sample of Twitter users that is representative for the Canadian population. The labeled dataset contains only 621 users, and with that limited number of training instances we extracted features for classical machine learning and achieved an F1-score of 0.922 using linguistic and emotion features. In addition, we fine-tuned a Sentence Pair Classification BERT model and achieved 92.6 F1-score. The classical machine learning trained model was applied on Canadian population representative dataset. The geographic and demographic patterns of suicide ideation correlate with the suicide statistics reported by Statistics Canada for 2015.
 

Index Terms—natural language processing, suicide, mental health, social media, deep learning

Cite: Ruba Skaik and Diana Inkpen, "Suicide Ideation Estimators within Canadian Provinces Using Machine Learning Tools on Social Media Text," Journal of Advances in Information Technology, Vol. 12, No. 4, pp. 357-362, November 2021. doi: 10.12720/jait.12.4.357-362

Copyright © 2021 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.