Predicting Mobile Application Ratings Using Artificial Neural Network

摘要

This paper presents and evaluates neural network models for predicting the ratings of mobile applications. It is shown that the information available to the developer at the time of releasing the application alone, can be used to make the predictions. The results of evaluation of 50,000 applications, available on a popular application store, is reported. It is seen that the prediction accuracy varies from 58.15% to 93.5% for different cases.