Privacy in Recommender Systems

Jeckmans, Arjan; Beye, Michael; Erkin, Zekeriya; Hartel, Pieter H.; Lagendijk, Reginald L.; Tang, Qiang

doi:10.1007/978-1-4471-4555-4_12

Cited by 89 publications

(53 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As similar privacy concern for the users' sensitive context data in [5], we consider a adversary model as follows: (1) Malicious third party who can gain access to the recommendation outputs and own some side information such as location about some users. The goal of this malicious third part is to deduce a particular user's features by observing the recommendation outputs.…”

Section: Adversary Model and Design Goalsmentioning

confidence: 99%

See 1 more Smart Citation

Differentially Private Online Learning for Cloud-Based Video Recommendation With Multimedia Big Data in Social Networks

Zhou

et al. 2016

IEEE Trans. Multimedia

View full text Add to dashboard Cite

Abstract-With the rapid growth in multimedia services and the enormous offers of video contents in online social networks, users have difficulty in obtaining their interests. Therefore, various personalized recommendation systems have been proposed. However, they ignore that the accelerated proliferation of social media data has led to the big data era, which has greatly impeded the process of video recommendation. In addition, none of them has considered both the privacy of users' contexts (e,g., social status, ages and hobbies) and video service vendors' repositories, which are extremely sensitive and of significant commercial value. To handle the problems, we propose a cloud-assisted differentially private video recommendation system based on distributed online learning. In our framework, service vendors are modeled as distributed cooperative learners, recommending videos according to user's context, while simultaneously adapting the video-selection strategy based on user-click feedback to maximize total user clicks (reward). Considering the sparsity and heterogeneity of big social media data, we also propose a novel geometric differentially private model, which can greatly reduce the performance (recommendation accuracy) loss. Our simulation shows the proposed algorithms outperform other existing methods and keep a delicate balance between computing accuracy and privacy preserving level.

show abstract

Section: Adversary Model and Design Goalsmentioning

confidence: 99%

“…On the one hand, as declared in [5], user's sensitive context information may be exposed by the recommendation results. Intuitively, the more detailed the information related to the user is, the more accurate the recommendations for the user are.…”

mentioning

confidence: 99%

Differentially Private Online Learning for Cloud-Based Video Recommendation With Multimedia Big Data in Social Networks

Zhou

et al. 2016

IEEE Trans. Multimedia

View full text Add to dashboard Cite

show abstract

“…Firstly, using collaborative filtering requires to store every user's privacy preferences somewhere on a server and protection of users' preferences is complex [22,23]. With a content-based recommender system, user's preferences are stored locally and are not shared at all.…”

Section: Introduction To Decision Support Systemmentioning

confidence: 99%

A new approach for managing Android permissions: learning users’ preferences

Oglaza

Laborde

Zaraté

et al. 2017

EURASIP J. on Info. Security

View full text Add to dashboard Cite

Today, permissions management solutions on mobile devices employ Identity Based Access Control (IBAC) models. If this approach was suitable when people had only a few games (like Snake or Tetris) installed on their mobile phones, the current situation is different. A survey from Google in 2013 showed that, on average, french users have installed 32 applications on their Android smartphones. As a result, these users must manage hundreds of permissions to protect their privacy. Scalability of IBAC is a well-known issue and many more advanced access control models have introduced abstractions to cope with this problem. However, such models are more complex to handle by non-technical users. Thus, we present a permission management system for Android devices that (1) learns users' privacy preferences with a novel learning algorithm, (2) proposes them abstract authorization rules, and (3) provides advanced features to manage these high-level rules. Our learning algorithm is compared to two other well-known approaches to show its efficiency. Finally, we prove this whole approach is more efficient than current permission management system by comparing it to Privacy Guard Manager.

show abstract

“…The main aim of ecommerce sites is to increase revenues, customer satisfaction and retention. For users, Recommender Systems lower the transaction costs of finding and selecting items, and generally improve decision quality [3], [4].…”

Section: Introductionmentioning

confidence: 99%

A Survey of Context-Aware Healthcare Recommender Systems

2017

IJSR

View full text Add to dashboard Cite

Abstract:The advent of social web has transformed users from sheer consumers to ardent producers of information. The rapid growth coupled with the reliance on the internet for information ranging from e-commerce, e-government and social networks has led to infobesity. Recommender Systems have become essential tool to aid users to overcome the problem of information overload and provide personalized relevant suggestions. The use of Recommender Systems have been successful in both industry and academia. Recommender systems have also found their way into the Healthcare domain with varied applications.Despite the growing practical demand, a few challenges are hindering the adoption ofRecommender Systems in the Healthcare.The purpose of this study is to understand the trend of Recommender Systems applications inHealthcareby examining the published literature, and to provide practitioners and researchers with insight and future direction.We conducted a focused literature review to identify research papers on Recommender Systems in the Healthcare in the last ten years. The key findings of this survey is that the application of Recommender Systems inHealthcare is evolving. The incorporation of contextual information is limited although it has been suggested as a key ingredient to improving the quality of the recommendations and the accuracy of the predications.It is important to point out that aside from the common filtering problems of sparsity, cold start and scalability, a socio-technical issue of privacy, security and trust is emerging.Collaborative Filtering technique is predominantly in use in Healthcare Recommender Systems. Incorporation of contextual information is limited.Privacy, security and trust concerns may hinder the widespread use and acceptance of the Healthcare Recommender Systems.

show abstract

Privacy in Recommender Systems

Cited by 89 publications

References 37 publications

Differentially Private Online Learning for Cloud-Based Video Recommendation With Multimedia Big Data in Social Networks

Differentially Private Online Learning for Cloud-Based Video Recommendation With Multimedia Big Data in Social Networks

A new approach for managing Android permissions: learning users’ preferences

A Survey of Context-Aware Healthcare Recommender Systems

Contact Info

Product

Resources

About