Suicide Risk Assessment with Multi-level Dual-Context Language and

Abstract

Mental health predictive systems typically model language as if from a single context (e.g. Twitter posts, status updates, or forum posts) and often limited to a single level of analysis (e.g. either the message-level or userlevel). Here, we bring these pieces together to explore the use of open-vocabulary (BERT embeddings, topics) and theoretical features (emotional expression lexica, personality) for the task of suicide risk assessment on support forums (the CLPsych-2019 Shared Task). We used dual context based approaches (modeling content from suicide forums separate from other content), built over both traditional ML models as well as a novel dual RNN architecture with user-factor adaptation. We find that while affect from the suicide context distinguishes with no-risk from those with "anyrisk", personality factors from the non-suicide contexts provide distinction of the levels of risk: low, medium, and high risk. Within the shared task, our dual-context approach (listed as SBU-HLAB in the official results) achieved state-of-the-art performance predicting suicide risk using a combination of suicide-context and non-suicide posts (Task B), achieving an F1 score of 0.50 over hidden test set labels.

Natural Language Analysis and the Psychology of Verbal Behavior: The Past, Present, and Future States of the Field

“…Due to their novelty, contextual embeddings have yet to see serious action in psychological research aside from classification benchmarking (e.g., Matero et al, 2019). However, we believe that the entire discipline of psychological language analysis is poised for a quantum shift.…”

A Time-Aware Transformer Based Model for Suicide Ideation Detection on Social Media

“…Contextual Methods: The best performing model, the DualContextBERT (Matero et al, 2019), at CLPsych 2019 for suicidal estimation exemplifies the utility of temporal context. The DualCon-textBERT models post embeddings sequentially via an RNN.…”

“…However, suicide ideation is complex, and often, individual posts may not be sufficient to assess a user's suicide risk, even for humans (Sisask et al, 2008;. Figure 1 illustrates how features such as historical posts (Matero et al, 2019) can add context for analyzing a user's online behavior over time (Van Heeringen and Marušic, 2003) to better ascertain suicide risk. Despite the success of usercentric contextual models (Flek, 2020) for suicide ideation detection, they have two major limitations.…”

Untitled

“…Team A B C Affective Computing 7 7 ASU (Ambalavanan et al, 2019) 2 5 CAMH † 5 2 2 Chen et al 20194 CLaC (Mohammadi et al, 2019) 1 5 1 CMU (Allen et al, 2019) 8 IBM data science (Morales et al, 2019) 12 10 4 IDLab (Bitew et al, 2019) 4 JXUFE † 9 8 SBU-HLAB (Matero et al, 2019) 3 1 3 TsuiLab (Ruiz et al, 2019) 3 TTU (Iserman et al, 2019) 6 8 UniOvi-WESO (Hevia et al, 2019) 10 11 uOttawa † 9 7 USI-UPF (Ríssola et al, 2019) 11 6 6…”