F
or this academic project, my classmate Nivethan and I wanted to create a basic movie recommender system. The goal was to create an application where a user could enter a movie title, and the system would return a list of 5 movies that the user might also enjoy, based on one of two parameters: keywords similarity or ratings similarity. The ratings similarity measure utilizes a K-nearest-neighbor algorithm to find similar movies based on similar user ratings, while the keywords similarity measure identifies movies based on frequency of similar keywords.
The goal of this project was to gain hands-on experience with implementing basic machine learning techniques into a usable standalone application. A number of data mining techniques were used to implement a working recommendation system, and although it works as intended, the system definitely has room for improvement. However, this project serves as a good starting point from which we can keep on working to improve our application.
The source code for the project is viewable on my GitHub page. Our final project report can be found here.
