We created a hack that can search through public twitter timeline histories of many people and determine whether they are at a risk of self harm or depression using personality profiling and sentiment analysis.

Organizations such as the local police or mental health support groups would be able to keep a close eye on those who are not in a good state of mind or having a rough time in life. Often people will express their feelings on social media due to the feeling on semi-anonymity and the fact that they can hide behind a screen, so it is possible a lot of people may be more transparent about heavy issues.

To connect our backend to our frontend, we took full advantage of the simplicity and utility of stdlib to create numerous functions that we used at various points to perform simple tasks such as scraping a twitter timeline for texts, sending a direct message to a specific user and one to interact with the Watson sentiment/personality analysis api. In addition, we have a website set up where an administrator would be able to view information gathered.

The next step would be setting up an automated bot farm that runs this project amongst relevant users. For example, a University mental support group would run it amongst the followers of their official Twitter account. It could also implement intelligent chat AI so that people can continue to talk and ask it for help even when there is nobody available in person.