PI Emiliano Albanese
The project is sponsored by the Zurich Foundation.
Worldwide, 800,000 people commit suicide every year. That is one person every 40 seconds, making suicide a public health priority. A key aspect of prevention is addressing suicidal thoughts before they become an accomplished act. However, these thoughts are not often expressed, and are thus difficult to detect and measure. With more than half of the global population (59%) now having access to the internet and internet users spending an average of 6,5 hours online each day2, social media are also becoming ubiquitous across sociodemographic groups (Perrin 2015). Social media are profoundly reshaping social life and social interactions, and impacting interindividual communication. Social media are also increasingly used to express psychological distress and its relationship with subjectively perceived and/or objective stressful circumstances that may contribute to suicidal ideation and suicide. The study of communication through social media is a promising strategy to capture suicidal thoughts, which would otherwise remain undetected.
RQ1: Is it possible to detect SI in tweets using our analysis approach and to which degree?
RQ2: How are SI tweets distributed in regards to space, time, and sociodemographic characteristics of the twitter user and the space within which the tweets are published?
RQ3: What are the emotions most often contained in SI tweets and do they differ in frequency and valence to the emotions contained in non-SI tweets? Is stress more common in SI tweets compared with non-SI tweets?
RQ4: Is there a significant regional correlation between SI incidence in Twitter messages and actual suicide rates? Is there an underlying spatial logic which can help explain a detected trend?