Human Computation Game for Enhancing Movie Box Office Prediction Model

Abstract

Decisions made in movie production are difficult to reverse and will possibly affect marketability of the movie. Movies are perceived differently by each group of audiences. To improve marketability, the movie must have some aspects to attract audiences. Previous study has proposed various methods for box office prediction but none has considered movie production as a continuous process nor using a game as an approach. In this study, we propose a human computation game, more suitable to determine subjective factors. Players have to invest in movie projects with limited information. The game is used in combination with mathematical model to enhance its performance. The combination model can continuously evaluate movie’s potential for the duration of movie production. We intend to explore how to combine both methods together and benefits of combination model.