Comparison of machine learning algorithms used to catalog Google Appstore

Abstract

Background: Google Play Store is a popular Android app store where users’ reviews and ratings provide valuable insights. As part of application development, clients and app designers have a significant impact on the market. Accurately predicting market trends is critical to the success of applications, and this is where information mining comes in. By evaluating various factors such as application name, pricing, reviews, and category, we can predict which types of apps are most likely to be successful. Methods: To do this, we can use machine learning techniques that help us analyze data from diverse metrics and identify relationships. In this work, five different types of existing machine learning computation techniques like decision tree algorithm (DTA), support vector machine (SVM), random forest (RF), stochastic gradient boosting (SGB) and gated recurrent unit (GRU) were used to show a comprehensive comparison between the models to describe and forecast the structure of Android Market applications. Through these methods, we can extract a wealth of information to make wise decisions and present data more effectively. Results: Our findings have shown that the GRU-based technique performance accuracy of our predictions is high, with an average of 89.4465%, which can be best visualized as an unusual forest picture better than the four baseline algorithms. Among all the five techniques, GRU classifies 5,616 apps and provides the most precise results for both users and developers, whereas the other classifiers only classified 3,744 apps. Conclusions: According to the result analysis, we can conclude that all five machine learning algorithms were capable of analysing the android market, GRU outperforms in terms of accuracy.