Movie Recommendation Program
Utilized C++ to read 3 data files from IMDB
- Title.akas.tsv.gz: used to find movies that were in the English language
- Title.basics.tsv.gz: used to find the movies titles, genres, and year of release
- Title.crew.tsv.gz: used to find the movie's director
All these components were pulled to create the movie itself, which was represented by a node.
We used an adjacency list implementation of a graph to link similar movies together. We calculated how similar the movies are by implementing our own weight function. Our weight was 1/(Similarrity Score) because the more similar the two movies, the closer they should be. The similarity score was increased by 2.5 for every genre that matched, increased by 1 if the director was the same, increased by 1.0 if there was explicit language, and increased by 1.0 if the start years of the two movies were within 5 years.
We also implemented a breadth-first search and depth-first search to offer the user two different sets of movies they could watch given an initial title from the user.
Our demonstration youtube video: https://youtu.be/N-vT5hFnFzc
Our documentation document: https://docs.google.com/document/d/15PXHhbESUSJbPOADwFNM8dlTRsYGWaA7kPrqL5VTQxk/edit?usp=sharing