Machine Learning on Box Office Prediction and the Impact of Protagonist Gender

Aug 5, 2026

Cinema’s influence on societies makes reliable box office prediction essential for reducing risks in future film investments. Accurate models support better resource allocation and improve the likelihood of backing successful films. The study addresses the commonly held assumption that lead protagonist gender affects commercial performance, where female-led films are often viewed as less likely to succeed and women remain underrepresented. Using a 2012–2019 and 2022–2023 dataset with inflation adjustment, it builds and evaluates machine learning predictors, finding Random Forest achieves strong match accuracy. Results show gender is statistically significant yet minimally impacts outcomes compared with budget, director ranking, star power, and social media ratings, guiding more evidence-based decision-making.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages47

Reading Time1 h 58 min

AccessFree to download

1. Introduction

2. Literature Review

2.1. Box office prediction

2.2. Gender representation and its impact on box office

3. Methodology

3.1. Data collection and integration

3.2. Feature definitions

3.3. Feature selection

3.4. Data reduction

3.5. Evaluation

MGI Master’s Degree Program in Information Management MACHINE LEARNING ON BOX OFFICE PREDICTION AND THE IMPACT OF PROTAGONIST GENDER José Ricardo Salas Pástor Master Thesis presented as partial requirement for obtaining the master’s degree in information management. NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa Machine Learning on Box Office Prediction and the Impact of Protagonist Gender by José Ricardo Salas Pástor Master Thesis presented as partial requirement for obtaining the master’s degree in Information Management, with a specialization in Knowledge Management and Business Intelligence Supervised by Carina Albuquerque, PhD., NOVA Information Management School December 2nd, 2024 STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Rules of Conduct and Code of Honor from the NOVA Information Management School. Lisbon, December 2nd, 2024. ACKNOWLEDGEMENTS I would like to express my deepest gratitude to my family and friends for their constant belief in me. Your support, especially in moments when time felt fleeting, has been my driving force throughout this journey. A special thanks to my supervisor, Professor Carina Albuquerque, for her guidance and insightful feedback. It was not an easy journey, and I am very grateful for the continuous support. I am forever indebted to my parents for their unwavering encouragement and for always pushing me to complete what I start. Your love and support made this work possible. I would also like to thank my professors at NOVA IMS, as well as the institution itself, for providing me with a rich learning environment and the resources necessary to complete this project. To my friends and classmates at NOVA IMS—Alisson, Camila, Cristina, María José, and Maytte—thankyou for being part of this incredible adventure. Moving to Lisbon together to start our master’s was an experience I’ll always cherish, and your companionship along the way made all the difference. Lastly, to all those who contributed to my growth through discussions, challenges, and shared moments of motivation: thank you. Your support has been invaluable and will never be forgotten. ABSTRACT The impact of cinema on societies, cultures, and social change is undeniable, and box office prediction plays a crucial role in minimizing risks associated with future film investments. Accurate forecasting models enable better resource allocation, improving the chances of supporting successful films. Historically, it has been assumed that the gender of the lead significantly influences a film's box office success, with female-led films often perceived as less likely to perform well commercially. This perception contributes to the underrepresentation of women in lead roles and persistent gender pay disparities in the film industry. To address these challenges and improve forecasting accuracy, this study develops a machine learning model for accurate box office prediction, analyzing the impact of various features on success, using a dataset of films released between 2012-2019 and 2022-2023. Unlike previous research, this dataset includes post-pandemic years and adjusts for inflation, providing a more up-to-date and comprehensive analysis. After evaluating multiple models, Random Forest emerged as the most effective, achieving 48.82% accuracy for exact matches and 80. 2% for matches within one revenue category. Additionally, a comprehensive analysis of gender’s impact is conducted by combining machine learning techniques with statistical tests, movin

FAQs

What is the main goal of the study on box office prediction?

To develop a machine learning model that predicts box office performance while analyzing how different features—including protagonist gender—relate to success.

How does the dataset used in this research improve over prior work?

It includes post-pandemic release years and adjusts revenues for inflation, enabling a more up-to-date and comprehensive analysis.

Which model performs best, and what accuracy levels are reported?

Random Forest performs best, reaching 48.82% accuracy for exact revenue matches and 80.2% for matches within a single revenue category.

Aug 5, 2026

Upload by River Wang