An empirical study of the use of multi-dimensional contexts for collaborative-filtering-based service recommendations in IoT environments

Abstract

Collaborative filtering (CF) based recommendation techniques involve the use of feedback information such as users’ ratings on items to predict their preferences on new items. To recommend services in Internet of Things (IoT) environments by utilizing CF based techniques, it is however essential to take into account multi-dimensional context information such as temporal, social, and spatial context to deal with the dynamism characteristic of IoT environments. In this paper, we propose a user-service matrix model to represent the contextual dependency between users and services and to analyze the feasibility of using multi-dimensional context information with regard to the accuracy of service recommendations. We conducted an experiment to demonstrate our approach using datasets collected from practical IoT testbed environments in which various smart devices are installed.