Achieving private SVD-based recommendations on inconsistently masked data

Yakut I., Polat H.

1st International Conference on Security of Information and Networks, SIN 2007, Gazimagusa, Turkey, 7 - 10 May 2007, pp.172-176 identifier

  • Publication Type: Conference Paper / Full Text
  • City: Gazimagusa
  • Country: Turkey
  • Page Numbers: pp.172-176
  • Anadolu University Affiliated: Yes


Users' concerns about private data might be different and they want various privacy levels. Therefore, they might decide to mask their data differently to achieve required privacy levels. Providing collaborative filtering (CF) services on inconsistently masked data is challenging. In this paper, we discuss how to achieve singular value decomposition (SVD)-based CF services on inconsistently masked data. © 2008 Atilla Elçi.