In the era of social media, identifying individuals with suicidal ideation has become increasingly important for crisis intervention and suicide prevention organizations. The researchers in this study sought to contribute to this field by using sentiment analysis to identify suicidal ideation in publicly available counseling transcripts. The team used the Linguistic Inquiry and Word Count (LIWC) Receptiviti API to apply sentiment analysis to a corpus of counseling transcripts categorized by symptom, then performed experiments to predict if a transcript described a suicidal patient based on the sentiment analysis data.
The researchers found that their initial results were promising, indicating that their methodology could be used in conjunction with analyzing social media text to improve technology-based suicide prevention efforts. Overall, this study highlights the potential for technology-based solutions to support crisis intervention and suicide prevention efforts.
Read the research: Automatic quantification of the veracity of suicidal ideation in counseling transcripts
