Suicide Ideation Detection via Social and Temporal User Representations using Hyperbolic Learning

Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2176–2190 June 6–11, 2021. ©2021 Association for Computational Linguistics 2176 Suicide Ideation Detection via Social and Temporal User Representations using Hyperbolic Learning Ramit Sawhney†‡∗and Harshit Joshi¶∗and Rajiv Ratn Shah‡ and Lucie Flek† † Conversational AI and Social Analytics (CAISA) Lab Department of Mathematics and Computer Science, University of Marburg http://caisa-lab.github.io ‡ Multimodal Digital Media Analysis (MIDAS) Lab Department of Computer Science and Engineering, IIIT Delhi http://midas.iiitd.edu.in ¶ Cluster Innovation Centre, University of Delhi Abstract Recent psychological studies indicate that in- dividuals exhibiting suicidal ideation increas- ingly turn to social media rather than mental health practitioners. Personally contextualiz- ing the buildup of such ideation is critical for accurate identification of users at risk. In this work, we propose a framework jointly lever- aging a user’s emotional history and social in- formation from a user’s neighborhood in a net- work to contextualize the interpretation of the latest tweet of a user on Twitter. Reflecting upon the scale-free nature of social network re- lationships, we propose the use of Hyperbolic Graph Convolution Networks, in combination with the Hawkes process to learn the histor- ical emotional spectrum of a user in a time- sensitive manner. Our system significantly out- performs state-of-the-art methods on this task, showing the benefits of both socially and per- sonally contextualized representations. 1 Introduction Every 40 seconds, a person dies by suicide (Roth et al., 2018). Despite the success of psychoclin- ical methods, such as the Suicide Probability Scale (Bagge and Osman, 1998) and Suicide Ideation Questionnaire (wa Fu et al., 2007), the sui- cide rate in the U.S. has risen by 35% in the last 20 years (Hedegaard et al., 2020). While these meth- ods are professional (Pestian et al., 2017), they have limited efficacy and may even impact participants negatively (Harris and Goh, 2017). Their limita- tions include barriers such as social stigma (Crisp et al., 2000), low literacy (Batterham et al., 2013), low motivation to seek hel (Essau, 2005), and fi- nances (Czyz et al., 2013). Tragically, 80% of patients do not undergo clinical treatment, and 60% of those who died by suicide denied having suicidal thoughts to practitioners (McHugh et al., 2019). ∗ Both authors contributed equally Contrarily, people turn to social media to express suicidal thoughts (Luxton et al., 2012; Coppersmith et al., 2014; Robinson et al., 2016), with 8 of 10 people disclosing their suicidal plans (Golden et al., 2009). Consequently, a growing body of work has shown that natural language processing can comple- ment social media analysis to identify risk markers in online user behavior to aid suicide risk assess- ment (McCarthy, 2010; De Choudhury et al., 2016; Reger et al., 2020; Shing et al., 2018). However, analyzing individual user posts is not always suffi- cient to infer user’s mental state and the associated suicide risk (Harris, 2010; Sisask et al., 2008). Studies suggest that suicide can be influenced by social factors (Masuda …

Content truncated to 500 words