According to MIT, "it is estimated that 17 percent of the U.S. population (between 5-12 percent of men and 10-20 percent of women) will suffer from a major depressive episode at least once in their lifetime." Fortunately, however, with strong support and reassurance, many such cases of poor mental health are issues that can be treated. We wanted to create software that would help to keep an eye on its user's mental health, offering calming/upbringing resources or even notifying the user's loved ones if the user's mental health seems to be getting worrisome. We wanted to create an app that would make an impact in the way those affected by mental illnesses received help by streamlining the process of finding support.

What it does

Using Google's Cloud Natural Language, the extension processes content on the pages the user visits, scores the words using a sentiment analysis process, and then processes a user mood score for how they're feeling. If the user mood score falls under a certain threshold, the extension will offer the user calming resources in efforts to help boost his/her mood.

How we built it

We used HTML, CSS, JS, Bootstrap, jQuery, Twilio's API, and Node.js

Challenges we ran into

Getting the styles to render correctly, brainstorming features for the extension.

Accomplishments that we're proud of

We built something presentable in 36 hours! In addition, using Google's Cloud Natural Language API was a great challenge that we were glad to overcome.

What we learned

We learned how to use JavaScript, Node.js, Bootstrap, Google's Cloud Natural Language and Twilio's APIs.

What's next for Reassuron

We're planning to use sentiment analysis to analyze user input and have our attachment react to sentiment scores produced by the extension. We want to create functionality for the application to offer to send a message to the user's contacts when the negative mood threshold is passed, so the user is encouraged to speak his or her feelings with his or her loved ones. Other possible avenues include sending automated text messages when negativity reaches a certain threshold, or evaluating whether the user could be a threat to others, as is often detectable from the browsing history of some school shooters, and alerting the loved one's contacts. We also have thought about the possibility of utilizing machine learning to create a more and more advanced method of detecting mental illness/depression.

Share this project: