We discover patterns related to depression in the social graph of an online community of approximately 20,000 lesbian, gay, and bisexual, transgender, and questioning youth. With survey data on fewer than two hundred community members and the network graph of the entire community (which is completely anonymous except for the survey responses), we detected statistically significant correlations between a number of graph properties and those TrevorSpace users showing a higher likelihood of depression, according to the Patient Healthcare Questionnaire-9, a standard instrument for estimating depression. Our results suggest that those who are less depressed are more deeply integrated into the social fabric of TrevorSpace than those who are more depressed. Our techniques may apply to other hard-to-reach online communities, like gay men on Facebook, where obtaining detailed information about individuals is difficult or expensive, but obtaining the social graph is not.
Tharindu Cyril Weerasooriya, Alexander Ororbia, Raj Bhensadadia, Ashiqur KhudaBukhsh, Christopher Homan. Findings of the Association for Computational Linguistics: ACL 2023. 2023.
19 November 2020 in Journal of Medical Internet Research
Background: Social media is a rich, virtually untapped source of data on the dynamics of intimate partner violence, one that is both global in scale and intimate in detail. Objective: The aim of this study is to use machine learning and other computational methods to analyze social media data for the reasons victims give for staying in or leaving abusive relationships. Methods: Human annotation, part-of-speech tagging, and machine learning predictive models, including support vector machines, were used on a Twitter data set of 8767 #WhylStayed and #WhylLeft tweets each. Results: Our methods explored whether we can analyze micronarratives that include details about victims, abusers, and other stakeholders, the actions that constitute abuse, and how the stakeholders respond. Conclusions: Our findings are consistent across various machine learning methods, which correspond to observations in the clinical literature, and affirm the relevance of natural language processing and machine learning for exploring issues of societal importance in social media.
Tong Liu, Christopher Homan, Cecilia Ovesdotter Alm, Megan Lytle, Ann Marie White, Henry Kautz. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2016.