App mining
Peng Fei Yu, Ching‐man Au Yeung · 2014
In this poster, we present a new model for estimating the actual value of mobile apps to the users. The model assumes that users are implicitly evaluating the value of the apps in their smartphones when they choose to uninstall some apps. Our proposed method thus makes use of the install and uninstall log in a mobile app store to estimate the value of the apps. Experiments using data from a popular mobile app store show that our model is better in predicting the future download trend of the apps as well as the future uninstallation rate of the apps. We believe such model will be very useful in generating more credible and appropriate mobile app recommendations to users, or in generating features for machine learning systems in more complex prediction tasks.