Box-Office Analytics and Movie Recommender System Using Machine Learning Algorithms

Abstract

The film industry has boosted up with hundreds of movies during the pandemic. The film makers have tough and challenging time to produce movies which the audience are willing to watch. People all around the globe are binge watching their favorite movies and web – series in their pastime. Filmmakers are trying to boost up the quantity and quality of their projects because of high competition they are facing. Filmmakers need help to identify the kind of genres that people are willing to watch and also the film watchers with personalized recommendations.

A common platform is built for both film makers and viewers for predicting box office success and data analytics for analyzing trends in audience’s interests. The Film makers will be equipped with features such as predicting the box office success of their project using various parameters that the machine learning model has been trained on. They can understand and analyze the social media engagements and strategize accordingly. Also, the film makers can make use of the Location suggestion feature to get the suggested locations for canning various scenes of their movies / web – series. On the other hand, the general audience will be able to use recommendation systems based on parameters like genre, OTT Platform and many others. They can also vote for the genres they are willing to watch which would help the film makers in understanding the interests of the audience. In this paper, the model was tested using various machine learning algorithms such as linear regression, decision tree regression and gradient boosting regression. Gradient boosting regression has shown better results with r2 score of 0.8960.