Inspiration
Several members of our team have a lot of steam games and not so much time to sift through them all (too busy getting new games!). The more we thought about this, the more we realised this is an issue that many steam users face! The steam backlog problem is well known in the community, with sites like steamleft built to quantify this.
What it does
It is a web app that users can go to and login with their steam accounts to get a list of what games in their library they should play It uses machine learning based on collaborative filtering of user's data to generate a list of recommendations for games which users have not played that are in their library. The more people use it, the more data we get, the smarter it gets!
How we built it
Flask server, pyspark for machine learning, and some bootstrap for a quick and nice ui.
Challenges we ran into
Learning machine learning in 24 hours and integrating it into the web app. Polishing various aspects, such as load times, has been especially difficult.
Accomplishments that we're proud of
Getting a working machine learning model with reasonable results, and the fact that almost everyone we meet seems to be really into this idea.
What we learned
A whole lot of things.
What's next for Shrec
Memes, dreams, and steam.
Built With
- api
- bootstrap
- flask
- igdb
- machine-learning
- pyspark
- python
- steam
Log in or sign up for Devpost to join the conversation.