3248920
doi
10.1145/3297280.3297660
oai:zenodo.org:3248920
user-eu
Student Research Abstract: "Hard to Understand, Easy to Ignore": An Automated Approach to Predict Mobile App Permission Requests
Hatamian, Majid
Goethe University Frankfurt
info:eu-repo/semantics/openAccess
Creative Commons Attribution 4.0 International
https://creativecommons.org/licenses/by/4.0/legalcode
<p>In this paper, we propose a novel automated approach to predict the potential privacy sensitive permission requests by mobile apps. Based on machine learning (ML) and natural language processing (NLP) techniques, personal data access and collection practices mentioned in app privacy policy text are analyzed to predict the required permission requests. Further, the predicted list of permission requests is compared with the real permission requests to check whether there is any mismatch. We further propose user interface designs to map mobile app permission requests to understandable language definitions for the end user. The combination of these concepts provides users with special knowledge about data protection practice and behavior of apps based on the analysis of privacy policy text and permission declaration which are otherwise difficult to analyze. Initial results demonstrate the capability of our approach in prediction of app permission requests. Also, by exploiting our already proposed app behavior analyzer tool, we investigated the correlation between what mobile apps do in reality and what they promise in their privacy policy text resulting in a positive correlation.</p>
Zenodo
2019-04-01
info:eu-repo/semantics/conferencePaper
3248919
user-eu
Accepted
award_title=Privacy and Usability; award_number=675730; award_identifiers_scheme=url; award_identifiers_identifier=https://cordis.europa.eu/projects/675730; funder_id=00k4n6c32; funder_name=European Commission;
1579541904.002565
881440
md5:312b8df7d80de9eb18200e4fe7b3c957
https://zenodo.org/records/3248920/files/Hatamian-SAC19.pdf
public