As a producer, I've always wanted to give back to the community of listeners as I can understand the emotions and stories that music can create and tell. My friend and I plan on creating a system to diagnose and help those with depression through music listening history and wanted to ask a community for help.
Looking For: Last.fm users - To take a less than 5 minute survey, a standardized depression diagnosis/history questionnaire Survey Link: https://goo.gl/forms/ruaEhJFp8rKKQtcx1
Goal: Create a Machine Learning model that will use a person’s music listening history to predict if their depressed.
Why? Depression is a silent disease. Many that suffer may not even realize their situation and often not be in the right mindset to find help. In 2017 a research group made a model that scanned Instagram posts, and could identify depression 70% of the time, while doctors are successful just 42% of the time. If models can achieve an accuracy that high using Instagram, a form of social media where posts are often heavily filtered, imagine the potential of the data for something as deeply tied to emotions as music.
The more data we have, the more accurate our diagnosis will be.
Please, take a few minutes out of your day to help us help the world.
Submitted January 02, 2018 at 04:56PM by ZefyrMusic https://www.reddit.com/r/edmproduction/comments/7nowek/music_listening_history_to_predict_depression/?utm_source=ifttt