An agent-based approach for privacy-preserving recommender systems

  • Authors:
  • Richard Cissée;Sahin Albayrak

  • Affiliations:
  • DAI-Labor, TU Berlin, Berlin;DAI-Labor, TU Berlin, Berlin

  • Venue:
  • Proceedings of the 6th international joint conference on Autonomous agents and multiagent systems
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Recommender Systems are used in various domains to generate personalized information based on personal user data. The ability to preserve the privacy of all participants is an essential requirement of the underlying Information Filtering architectures, because the deployed Recommender Systems have to be accepted by privacy-aware users as well as information and service providers. Existing approaches neglect to address privacy in this multilateral way. We have developed an approach for privacy-preserving Recommender Systems based on Multiagent System technology which enables applications to generate recommendations via various filtering techniques while preserving the privacy of all participants. We describe the main modules of our solution as well as an application we have implemented based on this approach.