Inspiration: We were inspired by the rise in mental health issues in people around us at the University of Waterloo, and we heard that increased social media use had detrimental effects on mental health. We wanted to do something to counteract this trend.
What it does Wellness Angel takes the tweets posted by a user and compares them to a sentiment analysis on the tweet
How I built it We used twitters developer api and the tweepy api to take tweets from users. We trained a google cloud machine learning network to learn which tweets are negative and may indicate depression or other mental illnesses. We built a website using html, css, and javascript. The main purpose of the website is to pass the twitter apis a username, access the tweets, and get the analysis from the google ML network
Challenges I ran into Setting up the web server using LAMP was one challenge. Another was first trying to create our own AI. Another challenge was then getting meaningful information from Google's AI. Another issue was the web design because no one had done it before. No one had also done python so that was another difficulty
Accomplishments that I'm proud of The ability to implement machine learning for the first time in our lives is the first thing that comes to mind. Another accomplishment we are proud of is pairing the front-end web design with back end python code
What I learned We learnt about both the back and front ends of a website and how they communicate to come together and form one cohesive website. In addition, we explored different machine learning algorithms, how to use external APIs, how to create our own API, as well as new python and html/css syntax
What's next for Wellness Angel
Adding a functionality that allows wellness angel to reach out directly to the person’s close friends based on their twitter interactions and offer resources for helping someone with depression
Log in or sign up for Devpost to join the conversation.