Analysis and Prediction of Users' Emotional Tone in Reddit Mental Health Communities
Keywords:mental health, reddit, sentiment analysis, machine learning
The rise in the number of people afflicted by mental health problems has placed these disorders among the main public health problems worldwide. As a result, user activity in communities related to mental health in online social networks has spiked. Here we characterize user activity in mental health-related communities on Reddit and analyze how user interactions through posts and comments influence their emotional state. In particular, we investigate whether seeking help on these networks results in changes in the feelings expressed by users over time. We observe that authors of negative posts often write rosier comments after engaging in discussions, indicating that users' emotional state can improve due to social support. Our results show that users who start discussions in these communities writing posts expressing negative feelings, tend to write more positive comments at the end, that is, they present an improvement in their emotional tone. In addition, we propose predictive models to capture the variation of the emotional tone of these users. Our models could assist in interventions promoted by health care professionals to provide support to the mentally-ill.
Baba, T., Baba, K., and Ikeda, D. (2019). Detecting mental health illness using short comments. In International Conference on Advanced Information Networking and Applications, pages 265–271. Springer.
Ballona, P. R., Prado, P., de Almeida, J. M., and Marques-Neto, H. T. (2015). Analyzing the influence of pope’s tweets on his followers’ mood. In Proceedings of the 21st Brazilian Symposium on Multimedia and the Web, WebMedia ’15, page 93–100.
Barney, L. J., Griffiths, K. M., Jorm, A. F., and Christensen, H. (2006). Stigma about depression and its impact on help-seeking intentions. Australian & New Zealand Journal of Psychiatry, 40(1):51–54.
Bird, S., Klein, E., and Loper, E. (2009). Natural language processing with Python: analyzing text with the natural language toolkit. ”O’Reilly Media, Inc.”.
Blair, J. and Abdullah, S. (2018). Supporting constructive mental health discourse in social media. In PervasiveHealth ’18.
Cranford, J. A., Shrout, P. E., Iida, M., Rafaeli, E., Yip, T., and Bolger, N. (2006). A procedure for evaluating sensitivity to within-person change: Can mood measures in diary studies detect change reliably? Personality and Social Psychology Bulletin, 32(7):917–929.
Cunha, T., Weber, I., and Pappa, G. (2017). A warm welcome matters!: The link between social feedback and weight loss in/r/loseit. In WWW’17 Companion, pages 1063–1072. International World Wide Web Conferences Steering Committee.
De Choudhury, M. (2013). Role of social media in tackling challenges in mental health. In Proceedings of the 2nd international workshop on Sociallyaware multimedia, pages 49–52. ACM.
De Choudhury, M. and De, S. (2014). Mental health discourse on reddit: Self-disclosure, social support, and anonymity. In Proceedings of the International AAAI Conference on Web and Social Media. ICWSM.
Gardner, M. W. and Dorling, S. (1998). Artificial neural networks (the multilayer perceptron)—a review of applications in the atmospheric sciences. Atmospheric environment, 32(14-15):2627–2636.
Gkotsis, G., Oellrich, A., Hubbard, T., Dobson, R., Liakata, M., Velupillai, S., and Dutta, R. (2016). The language of mental health problems in social media. In Proceedings of the Third Workshop on Computational Lingusitics and Clinical Psychology.
Goldberg, Y. (2017). Neural network methods for natural language processing. Synthesis Lectures on Human Language Technologies, 10(1):1–309.
Gruda, D. and Hasan, S. (2019). Feeling anxious? perceiving anxiety in tweets using machine learning. Computers in Human Behavior, 98:245– 255.
Hutto, C. J. and Gilbert, E. (2014). Vader: A parsimonious rulebased model for sentiment analysis of social media text. In ICWSM.
Islam, M. R., Kabir, M. A., Ahmed, A., Kamal, A. R. M., Wang, H., and Ulhaq, A. (2018). Depression detection from social network data using machine learning techniques. Health information science and systems, 6(1):8.
Newman, M. E. (2003). The structure and function of complex networks. SIAM review, 45(2):167–256.
Pellert, M., Schweighofer, S., and Garcia, D. (2020). The individual dynamics of affective expression on social media. EPJ Data Science, 9(1).
Sahota, P. K. and Sankar, P. L. (2019). Bipolar disorder, genetic risk, and reproductive decision-making: A qualitative study of social media discussion boards. Qualitative health research.
Shen, J. H. and Rudzicz, F. (2017). Detecting anxiety through reddit. In Proceedings of the Fourth Workshop on Computational Linguistics and Clinical Psychology—From Linguistic Signal to Clinical Reality, pages 58–65.
Silveira, B., da Silva, A. P., and Murai, F. (2020). Modelos de previsão do tom emocional de usuarios em comunidades de sa ´ ude mental no reddit. In ´ Anais do IX Brazilian Workshop on Social Network Analysis and Mining, pages 13–24, Porto Alegre, RS, Brasil. SBC.
Silveira, B., da Silva, A. P. C., and Murai, F. (2018). Analise de comunidades de suporte a transtornos de saude mental do reddit. In ´ Anais do VII Brazilian Workshop on Social Network Analysis and Mining. SBC.
Silveira Fraga, B., Couto da Silva, A. P., and Murai, F. (2018). Online social networks in health care: A study of mental disorders on reddit. In 2018 IEEE/WIC/ACM International Conference on Web Intelligence (WI), pages 568–573.
Stone, A. A., Hedges, S. M., Neale, J. M., and Satin, M. S. (1985). Prospective and cross-sectional mood reports offer no evidence of a”blue monday”phenomenon. Journal of Personality and Social Psychology, 49(1):129.
Weerasinghe, J., Morales, K., and Greenstadt, R. (2019). “because... i was told... so much”: Linguistic indicators of mental health status on twitter. Proceedings on Privacy Enhancing Technologies, 2019(4):152–171.
Wolohan, J., Hiraga, M., Mukherjee, A., Sayyed, Z. A., and Millard, M. (2018). Detecting linguistic traces of depression in topic-restricted text: Attending to self-stigmatized depression with NLP. In Proceedings of the First International Workshop on Language Cognition and Computational Models, pages 11–21, Santa Fe, New Mexico, USA. Association for Computational Linguistics.
Wongkoblap, A., Vadillo, M. A., and Curcin, V. (2019). Modeling depression symptoms from social network data through multiple instance learning. AMIA Summits on Translational Science Proceedings, 2019:44.
How to Cite
Copyright (c) 2021 The authors
This work is licensed under a Creative Commons Attribution 4.0 International License.