Steam Similarity Recommender Find your next favorite game and learn WHY (student project)[P]

Short summary

A student built a Steam game recommender that explains why a game is recommended by breaking titles into fine-grained feature vectors (gameplay focus, vibe, music style) rather than broad tags. The system analyzed 2K reviews from 80K games through a ChatGPT-powered NLP pipeline to extract and group meaningful features into micro-tags. It's live at nextsteamgame.com with an open GitHub repo and advanced mode for custom feature weighting.

• •Built a Steam game recommender with explainable recommendations using fine-grained feature vectors instead of broad tags
• •Processed 80K games via ChatGPT-powered NLP pipeline to extract gameplay, aesthetic, and thematic features
• •Live demo at nextsteamgame.com; open-source on GitHub with PR opportunities and advanced tuning mode

Generated with AI, which can make mistakes.