In 2018 only in US and Canada was released 871 movies. But despite such several release films, viewers often leave the session dissatisfied because, looking at the trailer, they expect more or something completely different. Also, do not forget that everyday streaming services are developing and provide users with more and more high-quality content, pushing cinemas to the background.
Consumers need personalized content and we decided to provide it through machine learning. After the project was released, we were approached by one of the largest movie ticket booking service in order to implement the developed system on their website.
Recommended system Movieterra built on content-based filtering – recommendations are based on the description of the film and the relevant user data. The system adapts to user preferences in real time and provides relevant movies premieres, reaching the box office. The service does not require registration and allows you to instantly start a search. The user just needs to enter his favorite titles in the search bar and the algorithm will immediately give suggestions that he likes.
Among the technologies used are Python, Word2Vec, Gensim.
"Great project! The site is super slick. Thanks for doing this"
"We used this system on our website. Views and positive feedback increased. Users are satisfied. Thanks for the good work!"