Estimating the Box Office Success through Machine learning and Sentiment Analysis of Movie Feedback

The machine learning framework integrated with sentiment analysis, which uses data from social media sources to predict box office revenue and audience response. Real time data from Application programming interfaces such as Twitter and Instagram are gathered, filtered to exclude useless data, then subjected to sentiment analysis using the VADER and TextBlob libraries. The Random Forest and Linear Regression algorithms are applied to determine sentiment trends and box office revenue with relative maximum values of accuracy. The system illustrates various aspects of the sentiment which include the distribution of the sentiment, the density of the selected keywords, and the analysis of the trends with time. In this piece, it underlines the need to mix audience response data with other styles of modeling for a richer viewing of movie success. The features that can be added in future are real-time sentiment analysis, Deep Learning models to enhance the predictive nature of the algorithm.