I collect video games, mostly focusing on titles released in the 70s-90s, and there are a ton of awful games that I’ve picked up over the years. I’m currently working on a classification project that will help me find games that I haven’t played before that I’d actually enjoy.
I’m building a dataset from my existing collection that includes features such as gaming platform, developer, publisher, year of release, genre, and a description. I’m labeling each with a score from 1-10 which represents how much I personally enjoyed the game. Then I’ll feed in data from the Giant Bomb API (https://www.giantbomb.com/api/) to find games that are likely to score well for me.
Eventually I’d like to also add in some time series forecasting to help me determine when is the right time to buy a game, since prices tend to fluctuate quite a bit.