Box Office Performance using Movie Scripts using Kernel Based Approach

A methodology to predict box office performance of a movie at the point of green-lighting, when only its script and estimated production budget are available. Initially extracting three levels of textual features (genre and content, Dialogue & scene variable, and bag-of-words) from scripts using screenwriting domain knowledge, human input, and natural language processing techniques. These textual variables define a distance metric across scripts which is then used as an input for a kernel-based approach to assess box office performance. In the proposed system mainly considered hindi movie script and its genre content variable and a script's Dialogue & scene variable features and by use of natural language processing techniques like stop word removal, stemming algorithm and LDA the obtained textual data is reduced which is then used as an input to kernel based method, Here by Kernel method output is obtained which gives lowest MSE on compared with existing methods which shows that proposed methodology predicts box office revenues more accurately lower mean squared error (MSE)) compared to benchmark methods. Through a portfolio selection scenario, demonstrating the potential economic significance of proposed approach by showing that this method generates portfolio returns that are significantly higher than portfolios selected by a comps-based approach that mimics the current industrial practice.