VectorCinema
Semantic movie recommendations — describe a mood, get films. Sentence-BERT embeddings in Astra DB, 94% precision@10.
01
Search by mood, not title
Instead of keyword matching, VectorCinema embeds every film's plot and themes with Sentence-BERT and stores them in Astra DB, so you can type 'sci-fi films in space' and get meaningfully related results.
02
Recommendations you can see
Every film shows its nearest neighbours with similarity scores, calculated live from the vector index.
03
Results
94% precision@10Natural-language searchLive similarity scores
FAQ
Questions people ask
Why vectors?
Embeddings capture meaning, so two films with no shared keywords can still be found as similar.
NEXT CASE STUDYBunkToBrainsStudy platform for GGSIPU students Read it
