Google Play Store Analysis

Abstract

Google play store consists of millions of applications and several thousand apps are added on the play store every day. The competition is so fierce that it is really difficult for the developers to find out whether the app that is the product of his hard work is going to be successful or not. The main goal of the study is to create a tool that helps developers and organizations to understand the current trends in the Play Store and the user bias for applications and utilize the data for making optimal decisions regarding their upcoming app.The success of an app can be determined by factors like ratings, number of installs and reviews. In this studywe demonstrate using applied exploratory data analysis to discover correlaations among features of an app and direct contribution of features towards success of the app to predict which apps will succeed. Data from Google Play Store was used to train various supervised learning models for prediction of the rating of an app, the models being-Random Forest, Support Vector Machine and Linear Regression. The user reviews are used to learn input bias of users towards applications using natural language processing techniques. Data of reviews is used to train classification models to predict polarity and sentiment of reviews.