The Durkheim Project wants to help prevent suicides by implementing an artificial intelligence (AI) algorithm for Facebook, Twitter, and LinkedIn that identifies common words and phrases used by people contemplating suicide.
The project launched on July 2 and currently only targets combat veterans because of their disproportionately high suicide rates. The project is opt-in and requires an app to be installed on the users computers, iOS, and Android devices.
As users post messages they are uploaded to a medical database which then monitors those messages for troubling signs of potential suicide.
The database is tracked at the Geisel School of Medicine at Dartmouth University and data collected is never shared with third party members.
According to Chris Poulin, principal investigator on the Durkheim Project:
“The study we’ve begun with our research partners will build a rich knowledge base that eventually could enable timely interventions by mental health professionals. Facebook’s capability for outreach is unparalleled.”
At this time information collected by the Durkheim Project will not be used to stop potential suicides. The purpose for now is simply to collect data points. IN the
So far the study has discovered that 65 percent of users who have committed suicide have used specific keywords and phrases.
In the meantime the study so far has only showed a correlation between suicides and keywords/phrases and not a direct link.
Ultimately the Durkheim Project’s founders would like the platforms algorithms to be used on a more broad scale with the potential for third party notifications.
Do you think a computer program can really predict possible suicides via social networks?