Alambic: a privacy-preserving recommender system for electronic commerce

Aïmeur, Esma; Brassard, Gilles; Fernandez, José M.; Onana, Flavien Serge Mani

doi:10.1007/s10207-007-0049-3

Cited by 111 publications

(77 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Clustering Advertisers can use demographic filtering [21] to categorize subscribers based on their demographics and send advertisements accordingly. Since there are seven demographic features in our data model and total number of keywords pertaining to the demographic features is 42, the advertiser sends a total of 42 different (single keyword) queries to obtain the information about subscribers' demographics.…”

Section: Timing Results For Data Mining Algorithmsmentioning

confidence: 99%

See 2 more Smart Citations

A practical privacy-preserving targeted advertising scheme for IPTV users

Khayati

Örencik

Savaş

et al. 2015

Int. J. Inf. Secur.

View full text Add to dashboard Cite

In this work, we present a privacy-preserving scheme for targeted advertising via the Internet Protocol TV (IPTV). The scheme uses a communication model involving a collection of subscribers, a content provider (IPTV), advertisers and a semi-trusted server. To target potential customers, the advertiser can utilize not only demographic information of subscribers, but also their watching habits. The latter includes watching history, preferences for IPTV content and watching rate, which are periodically (e.g., weekly) published on a semi-trusted server (e.g., cloud server) along with anonymized demographics. Since the published data may leak sensitive information about subscribers, it is safeguarded using cryptographic techniques in addition to the anonymization of demographics. The techniques used by the advertiser, which can be manifested in its queries to the server, are considered (trade) secrets and therefore are protected as well. The server is oblivious to the published data and the queries of the advertiser as well as its own responses to these queries. Only a legitimate advertiser, endorsed with so-called trapdoors by the IPTV, can query the cloud server and access the query results. Even when some background information about users is available, query responses do not leak sensitive information about the IPTV users. The performance of the proposed scheme is evaluated with experiments, which show that the scheme is practical. The algorithms demonstrate both weak and strong scaling property and take advantage of high level of parallelism. The scheme can also be applied as a recommendation system.

show abstract

Section: Timing Results For Data Mining Algorithmsmentioning

confidence: 99%

“…, m σ ). Then the prediction for the interest of a subscriber s on the advertisement c can be calculated as [21]:…”

Section: Content-based Filteringmentioning

confidence: 99%

See 1 more Smart Citation

A practical privacy-preserving targeted advertising scheme for IPTV users

Khayati

Örencik

Savaş

et al. 2015

Int. J. Inf. Secur.

View full text Add to dashboard Cite

show abstract

“…The key components and concepts of the ALAMBIC architecture, described in detail in [2] and shown in Figure 1, are the following:…”

Section: The Alambic Architecturementioning

confidence: 99%

“…In earlier work [2,3], we have proposed techniques to fight on behalf of the customer against profile creation or dossier constitution-and ultimately against the looming threat of Big Brother [10]. Our approach is designed to help the customer buy products without compromising the privacy of his profile.…”

Section: Introductionmentioning

confidence: 99%

Experimental Demonstration of a Hybrid Privacy-Preserving Recommender System

Aïmeur

Brassard

Fernandez

et al. 2008

2008 Third International Conference on Availability, Reliability and Security

Self Cite

View full text Add to dashboard Cite

Recommender systems enable merchants to assist customers in finding products that best satisfy their needs. Unfortunately, current recommender systems suffer from various privacy-protection vulnerabilities. We report on the first experimental realization of a theoretical framework called ALAMBIC, which we had previously put forth to protect the privacy of customers and the commercial interests of merchants. Our system is a hybrid recommender that combines content-based, demographic and collaborative filtering techniques. The originality of our approach is to split customer data between the merchant and a semitrusted third party, so that neither can derive sensitive information from their share alone. Therefore, the system can only be subverted by a coalition between these two parties. Experimental results confirm that the performance and user-friendliness of the application need not suffer from the adoption of such privacy-protection solutions. Furthermore, user testing of our prototype show that users react positively to the privacy model proposed.

show abstract

OpenNym: Privacy preserving recommending via pseudonymous group authentication

Checco

Bracciale

Leith

et al. 2021

Security and Privacy

View full text Add to dashboard Cite

A user accessing an online recommender system typically has two choices: either agree to be uniquely identified and in return receive a personalized and rich experience, or try to use the service anonymously but receive a degraded non-personalized service. In this paper, we offer a third option to this "all or nothing" paradigm, namely use a web service with a public group identity, that we refer to as an OpenNym identity, which provides users with a degree of anonymity while still allowing useful personalization of the web service. Our approach can be implemented as a browser shim that is backward compatible with existing services and as an example, we demonstrate operation with the Movielens online service. We exploit the fact that users can often be clustered into groups having similar preferences and in this way, increased privacy need not come at the cost of degraded service. Indeed use of the OpenNym approach with Movielens improves personalization performance.

show abstract

Alambic: a privacy-preserving recommender system for electronic commerce

Cited by 111 publications

References 56 publications

A practical privacy-preserving targeted advertising scheme for IPTV users

A practical privacy-preserving targeted advertising scheme for IPTV users

Experimental Demonstration of a Hybrid Privacy-Preserving Recommender System

OpenNym: Privacy preserving recommending via pseudonymous group authentication

Contact Info

Product

Resources

About