ProtectMyPrivacy

Abstract

In this paper we present the design and implementation of ProtectMyPrivacy (PMP), a system for iOS devices to detect access to private data and protect users by substituting anonymized data in its place if users decide. We developed a novel crowdsourced recommendation engine driven by users who contribute their protection decisions, which provides app specific privacy recommendations. PMP has been in use for over nine months by 90,621 real users, and we present a detailed evaluation based on the data we collected for 225,685 unique apps. We show that access to the device identifer (48.4% of apps), location (13.2% of apps), address book (6.2% of apps) and music library (1.6% of apps) is indeed widespread in iOS. We show that based on the protection decisions contributed by our users we can recommend protection settings for over 97.1% of the 10,000 most popular apps. We show the effectiveness of our recommendation engine with users accepting 67.1% of all recommendations provide to them, thereby helping them make informed privacy choices. Finally, we show that as few as 1% of our users, classified as experts, make enough decisions to drive our crowdsourced privacy recommendation engine.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 183 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.