Inspiration: Paranoi affected one of my friends' life significantly. Delusion, hallucination and the feeling of inferiority drove her academic performance down. She was wary of her surroundings to the point that she was no longer feeling comfortable going out during the day. Yet it might be surprising to you that she did not feel the need for any professional help for over a year. No one could possibly convince her -- even me as her best friend. As I was consoling her over the phone for hours, I thought some objective tests should be there to get her supported.

What it does: this mental health checker provides a relatively reliable diagnosis with less than 30 questions. It clears the anxiety patients may face when seeking professional help.

How we build it We use the vector space model based on term weights and frequency combined with similarity functions to find the most likely mental disease based on the symptom evaluation from the user. After getting the user’s answers to the test, we retrieve professional articles and use query expansion to improve the overall information retrieval performance.

Challenges None of the members on the team has a psychology background, so for this topic, it was time-consuming for us to grasp the necessary background information and find and build a reliable and effective database at first.

What we learned: in the tech world, things do not go as expected a lot of the times.

What's next for Mental Health Checker: better UI design!!!

Built With

Share this project:

Updates