Predictive Modeling of App Ratings: A Review of Classical and Neural Network Methodologies

Abstract

The Google Play Store, a dominant platform for Android applications, generates vast amounts of data including app descriptions, user reviews, and ratings. Analyzing this data is crucial for developers to understand user preferences, improve app quality, and predict future ratings. This review paper synthesizes recent research on Play Store app analysis and rating prediction, focusing on the application of classical machine learning (ML) models and artificial neural networks (ANNs). We discuss various feature engineering techniques, model selection strategies, and performance evaluation metrics, highlighting the strengths and limitations of different approaches. This paper provides a comprehensive overview of the current state of research and identifies potential avenues for future exploration.