Apps Rating Classification on Play Store Using Gradient Boost Algorithm

Abstract

The increasing number of Android apps available on Google Play Store with the developers' advantages has attracted many Android apps developers' attention. To benefit from developing Android apps is to know the characteristics of high rated applications on the Google Play Store. This research will explore the features of Size, installs, reviews, types (free/paid), rating, Category, content rating, and Price on apps on Google Play Store to determine the characteristics of high rated apps. This research uses a random-forest classifier to identify the most significant features in high rated apps on Google Play Store. This research uses the Gradient Boost Algorithm to identify the most influential attributes in high rating apps on Google Play Store. To classify the high rated apps, writers use the Gradient Boost algorithm that performs better than Random Forest, K-NN, and Decision Tree algorithm with a 99.93% accuracy, 99.91% recall, 99.91% precision, and 0.062 Root Mean Squared Error.