Permission recommender system for Android

Ankur Shukla, Divya Vikash, Bharavi Mishra, Poonam Gera · 2017

There are 1.4 billion Android devices around the globe. With so many users and apps comes the question of privacy and security. Some permissions are critical to the operation of any app. A system is required which can tell whether an app X should be given a permission Y. Our basic idea is that an app belonging to a particular category should request permissions most common in that category. Around 130 users contributed to our dataset which had over 3964 records and 830 unique apps in 39 categories. A voting measure based on statistical mean value was used to decide the permissions to be given to particular app.

Read the paper · More papers on PaperTik