Abstract
In this paper, we propose a new privacy solution for the data used to train a recommender system, i.e., the user–item matrix. The user–item matrix contains implicit information, which can be inferred using a classifier, leading to potential privacy violations. Our solution, called Personalized Blurring (PerBlur), is a simple, yet effective, approach to adding and removing items from users’ profiles in order to generate an obfuscated user–item matrix. The novelty of PerBlur is personalization of the choice of items used for obfuscation to the individual user profiles. PerBlur is formulated within a user-oriented paradigm of recommender system data privacy that aims at making privacy solutions understandable, unobtrusive, and useful for the user. When obfuscated data is used for training, a recommender system algorithm is able to reach performance comparable to what is attained when it is trained on the original, unobfuscated data. At the same time, a classifier can no longer reliably use the obfuscated data to predict the gender of users, indicating that implicit gender information has been removed. In addition to introducing PerBlur, we make several key contributions. First, we propose an evaluation protocol that creates a fair environment to compare between different obfuscation conditions. Second, we carry out experiments that show that gender obfuscation impacts the fairness and diversity of recommender system results. In sum, our work establishes that a simple, transparent approach to gender obfuscation can protect user privacy while at the same time improving recommendation results for users by maintaining fairness and enhancing diversity.
Highlights
The data used to train a recommender system takes the form of a user–item matrix, where the columns represent items in the collection and the rows represent individual users
We focus on gender obfuscation, but Personalized Blurring (PerBlur) would be suited for protecting other sorts of information that can be inferred from user profiles
We show the interplay between user-profile obfuscation and fairness (Section 7) and diversity (Section 8) and demonstrate the potential of PerBlur to contribute in both cases
Summary
The data used to train a recommender system takes the form of a user–item matrix, where the columns represent items in the collection and the rows represent individual users. The user–item matrix does not explicitly contain specific user attributes such as gender Such information is implicit in each profile, since it can be predicted or inferred using machine learning, a classifier. As the user profiles collected and stored by online platforms increase in number and length, classifiers have a larger amount of data available for training and inference, and the privacy threat grows. To counter this threat, we need the right privacy solutions. Friedman, Boreli, and Sivaraman (2014) proposed ‘‘Privacy Canary’’, an interactive system that enables users to interact and control the privacy-utility trade-off of the recommender system to achieve a desired accuracy while maintaining privacy protection
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
