This is my first personal project, where I explored overlapping my interests in film, data, and coding!
MoodMatch is a mood-based movie recommendation site where you can input how you're feeling, or how you want to feel by the end of the movie, and get curated recommendations that match what you've entered.
What makes MoodMatch different is its focus on very specific and sometimes niche emotions, not just genre tags or oneworded feelings. Instead of typing "romance" or "comedy," you can type something like:
"You just finished your first year of college"
"It’s 3 AM and you’re feeling a little sad"
"It’s the night before Christmas and you’re feeling nostalgic"
and MoodMatch will interpret your input and return a short list of films that reflect those emotions and themes.
- Improve and expand mood mappings
- Refine keywords for better results
- Build UI
- Add features like movie posters and rating
- Additional NLP techs for better understanding of mood
--- to be updated ---
- summmary -
- challenges -
- features
Python 3, TMBD API, Hugging Face Transformers (sentence-transformers/all-MiniLM-L6-v2), Pytorch for embeddings and cosin similarity, Developed with VSCode