2022 EDSIG Proceedings: Abstract Presentation

Using Machine Learning to Predict Online Movie Rating Based on the Movie Characteristics: A Team Pattern Perspective

Qiunan Zhang
University of Memphis

Xihui Zhang
University of North Alabama

According to Film and Video Global Market Report 2021, the global film and video market reached to nearly $234.9 billion in 2020. The market is expected to reach $318.2 billion by 2025 and $410.6 billion by 2030. Nevertheless, the financial success of a movie is largely uncertain due to many pertinent factors. Understanding which factors influence the financial success of a movie is crucial for the industry. Online movie rating is always used to measure a movie’s success and has been studied in many IS studies. The factors influencing online movie ratings can be classified into two major types: internal and external. Internal factors, such as actors, actresses, director, and “stars,” influence movie rating and are related to the movie team. External factors, such as economic factors, the motivations for moviegoing, and the effects of advertising, are those factors which influence movie rating from outside environments. However, few studies focus on both the effect of individual and team characteristics on online movie ratings. Therefore, our research question is: How do team patterns (such as actors/actresses, experienced crew, and collaboration among team members) influence movie ratings?

Recently, the machine learning technologies, such as decision trees and neural networks, have been adopted in many studies and proved as one of the most potent methodologies in IS studies. This research will investigate the effects of movie team patterns on movie ratings by using machine learning technique - decision tree model. Decision tree is one of supervised machine learning methods and can model data that has nonlinear relationships between variables, and handle interactions between variables. In this research, we will also compare with OLS (ordinary least squares) regression and decision tree model.

This research will have two major contributions. First, the result could help the movie industry better understand how to form their team to get a higher online rating movie and thus achieve a better ROI. Second, this research will provide an example of applying machine learning in online users context analysis in the IS field.

Keywords: Online movie ratings, Team pattern, Machine learning, Decision tree

Thursday at 3:00 pm