Open AccessPredict App Rank on Google Play Using the Random Forest Method
Author(s) -
Abdul Khaliq,
Eko Hariyanto,
Supina Batubara
Publication year - 2021
Publication title -
international journal of research and review
Language(s) - English
Resource type - Journals
eISSN - 2454-2237
pISSN - 2349-9788
DOI - 10.52403/ijrr.20210955
Subject(s) - random forest , computer science , benchmark (surveying) , download , rank (graph theory) , app store , data mining , machine learning , world wide web , mathematics , geodesy , combinatorics , geography
Application developers and users are the keys to the market impact on application development. In application development, developers need to predict applications in the market accurately, accurate prediction results are very important in showing user ratings that affect the success of an application. Ratings are given by users to judge that the application is good or not. The higher the rating given by the user, it means that the user likes the application and can be a benchmark for other users to download the application. It is undeniable that there are so many apps available on the google play store, it is impossible for users to select one by one app on the google play store. Therefore, a rating prediction system is needed to determine the right application based on the rating given by the user to an application. Predictions will be made using the random forest algorithm as the method used to predict application ratings. This study using the Google Play Store dataset. This dataset has 10840 rows and 13 attributes. The results of this study can be seen from the use of the random forest algorithm with an average accuracy of 93.8%.Keywords: Google Play Store, Rating, Prediction, Random Forest.