A ubiquitous recommender system based on collaborative filtering and social networking data
Nikolaos Polatidis, Christos K. Georgiadis · International Journal of Intelligent Engineering Informatics · 2015
The use of mobile devices and the rapid growth of the internet and networking infrastructure has brought the necessity of using ubiquitous recommender systems. However, in mobile devices there are different factors that need to be considered in order to get more useful recommendations and increase the quality of the user experience. This paper gives an overview of the factors related to the quality and proposes a new hybrid recommendation model. The proposed model is based on collaborative filtering and social rating network data. Furthermore, it includes an approach to protect user privacy when context parameters are used, by transferring a subset of the users and ratings in the mobile device and applying the algorithm and context parameters locally. In addition, we recommend the use of classical user-based collaborative filtering, enhanced by the trust network, which is a method that performs better in terms of accuracy when compared with user-based collaborative filtering and trust-aware collaborative filtering. Our approach has been experimentally evaluated and is shown that is both practical and effective.