ANALYSIS OF GAMING APPROACHES TO STIMULATE USER ACTIVITY

Abstract

This paper examines the approaches used in gaming apps that influence mobile gaming app ratings. Using an open marketplace dataset, we analyzed the relationship between user ratings and price, app age, in-app purchases, and other characteristics. Exploratory data analysis methods were used to evaluate the key engagement factors presented in the dataset, and machine learning models, including linear regression, artificial neural networks, Random Forest, and XGBoost, were employed for data testing. Model quality was assessed using MAE, RMSE, and the R² coefficient of determination. The results showed no significant impact of app price upon download on user ratings. However, app age and the presence of in-app purchases demonstrate a more pronounced relationship with ratings. The resulting relationships are correlational, and further analysis of the cause-and-effect relationships is recommended. Based on the obtained results, recommendations were proposed for implementing in-app monetization, where the statistical correlation was more clearly revealed, as well as for more frequent update support. Furthermore, directions for further research are proposed for incorporating an expanded dataset and analyzing user behavior within the app. In the future, this work may help prioritize gamification factors when developing a mobile app to improve user experience.