Movie Recommendation System through Movie Poster using Deep Learning Technique
Harshali Desai · International Journal for Research in Applied Science and Engineering Technology · 2021
Movie recommendation system plays an important role for all media service providers like Netflix, Amazon Prime etc. to increase their business by providing user with appropriate list of movies from very broader lists.By helping user to choose movies of their interest, media service providers thus attract huge traffic which indeed helps them to increase revenue.Very often target audience opt to watch movie by just looking at the movie poster.Thus, users can immediately gain an idea about a particular movie from its movie posters.So, from movie posters one can easily guess the genres of the movie.In this paper we have proposed movie recommendation system which recommends the movies based on its genres which are predicted through its movie posters.Firstly, the movie posters are obtained from IMDB website by performing web scraping.These movie posters are then used to train the Convolution Neural Networks.Convolution Neural Networks then predicts the genres of the movie and then based on predicted genres movies are recommended to the end user based on their cosine similarity.The proposed model is than compared with the existing baseline model and found that our proposed model performs slightly better in terms of accuracy, f1-score and precision.