The Study And Implementation Of Automated Detection System In Android Abusive Push Notifications

Push notification is a significant module in mobile platforms,and has been widely used to deliver all sorts of information to users.It offers great convenience for app developers and enriches the experience of mobile users.However,this feature was recurrently abused recently in Android platform,such as delivering push notification advertisement,or even distributing malware.The abuse of push notifications has done great harm to the Android ecosystem,and yet,this problem has not been studied by our research community so far.To fill the void,this paper presents the first study to detect abusive push notifications in Android platform,and further characterize this phenomenon within multiple app markets through a large-scale experiment.We first provide a systematic taxonomy of Android push notifications and further identify the abusive ones.Then we design and implement an automated system to detect abusive push notifications in Android apps.Through the test policy of app queue,and by traversing the app based on viewtree,we are able to trigger and click on the push notifications in apps.By the instrumentation technology of the Android framework,we collect all the notification-relevant runtime information to flag the abusive ones.Our experimental results prove the effectiveness of the detection system.By applying the system to 20,000 Android apps from different app markets,we successfully yield over 1,000 abusive notifications.We also perform a fine-grained analysis on these notifications.Our study reveals that abusive push notifications are prevalent across different markets.Our community should focus more on the subject and take actions to mitigate this issue.