In this movie recommendation, the movies are recommended based on the content-based recommendation engine. In the project, we have a dataset of the movies. In that dataset, it contains director, keyword, genres, cast, etc. In content-based recommendation engine take input as user liked the movie. Then its analysis the content (which is dataset) like (genres, director, keyword, cast) of the input movie to find out other movies which have similar output. Then it ranks similar movies to its similarity score and recommends the movie to the user.
In the browser, if we search movie name it will give similar movies related to that input movie. This is done by three recommendation engine such as popular based recommendation engine, content-based recommendation engine, collaborative filtering-based recommendation engine.
The popular based recommendation engine is nothing but trending list on YouTube, Netflix, etc. They will keep on tracking the video views and update the on-trending list.
The content-based recommendation engine is taking the input from the user and give a similar rank list based on the content related to the input movie. In this project, the content-based recommendation engine is used.
The collaborative filtering-based recommendation engine is first search the similar user based on their activity and behaviour and if the first user and second user have seen the same movie and if the first user has seen the new movie but if the second user not. It recommends new movies to a second user and vice-versa.
Since the system uses the Scikit learning feature extraction text which converts a collection of a text document into matric token i.e. how many same words present in it. From that, it uses the pairwise to give the percentage related to the matric token.
So, it gives the similar movies ranks list related to the user movies.
Block Diagram Description:
This block diagram tells the details explanation of code. If the user enters the movie name in combine feature will give the name, genres, keyword, cast, director of the movie. Then feature text extraction to pairwise and last the rank list.
In this movie recommendation, the content-based recommendation engine is used. In this, the user gives input as a movie name. Then the content of the input movie is taken by combine feature (feature like keyword, cast, genres, director). From there the feature text extraction is done on the base of the combine feature in the dataset. Then the matrices pairwise are done on the base of feature text extraction. After that, the similar rank list in given based on the similarity score of the input movie and recommendation is done on the base of relevant to input movie.
- Scikit Learn
This content-based movie recommendation engine gives a good acquires on movie recommendation.