14 DAY TRIAL // With the number of teen suicides on the rise due to online bullying, having a way to anticipate such actions could result in a saved life. Considering how many people, teenagers and army veterans alike (two groups that are known to have a high suicide risk), take to social media to post their issues, some clues might be buried within their messages.
That’s what a Boston project is betting on. The Durkheim Project combines search technology with predictive analysis in order to estimate the suicide risk of an individual based on what they post on Facebook, Twitter or LinkedIn, Mashable reports.
The world’s largest social network, Facebook, is already collaborating with the Durkheim Project in gathering social media data from participating active duty military members and veterans in what they consider to be the first stage.
The tool analyzes daily status updates on Facebook, Twitter messages, as well as structure changes and keywords used on LinkedIn.
The project looks for certain keywords that can indicate a high risk for suicide, psychiatric illness or whether someone is fine.
The system seems to be quite smartly created and can even tell when someone uses some terms as a joke rather than actually express their intention to commit suicide.
Thus, lines such as “I’m going to kill myself if I miss my train,” can be detected and ignored, which means such phrases will not be correlated to suicide risk.
The Durkheim Project is seeking to get 100,000 individuals to opt in during the data collection period, which should help it become more accurate. While it’s true that it will most likely not save everyone, it could help make a difference and prevent such actions.
It remains to be seen how long it’s going to take to calibrate the system before it can be widely used.