Ziyuan Wang
Fun Project

Self

Movie + TV Recommender (up till 2024)

Input your preferences and get a recommendation from your top-rated movies + TV shows dataset. The recommender filters candidates and ranks them using rating, vote volume, and keyword relevance.

Workflow

  • Load CSV data and normalize fields (title, year, rating, votes, runtime, genre, certificate, description, stars).
  • Classify each title as Movie or TV Show using certificate values (for example, TV-MA means TV Show).
  • Apply your selected filters (type, genre, minimum rating, runtime cap, year range, keyword).
  • For the main recommendation: assign each remaining title a score using rating, vote volume, genre match, and keyword match, then rank descending.
  • Pick one title with weighted randomness from the top candidates so stronger titles are more likely, but results are not always identical.
  • Surprise Me uses uniform random selection from the chosen type pool (Movie, TV Show, or Any), without score weighting.

Dataset

Default dataset path: assets/data/top-rated-movies-2026.csv (movies and TV series supported)

Dataset not loaded yet.

Find a Movie / TV Show

Tip: if you have no preference for a filter, leave it empty (or keep it as "Any").

Load your dataset to start recommending.

Back to Projects