A Personalized Recommendation Framework for Social Internet of Things (SIoT)

Recommendation is inevitably crucial in human life, as almost every human daily activity involves decision and choice making from amongst various alternatives at our disposal. The use of computers has made automated decision making interesting, and beneficial in many areas of real-life activities, targeted at meeting different yet specific users' needs, such as in Social Internet of Things (SIoT). SIoT is defined as an emerging paradigm of IoT where intelligent devices are able to create social relationships among them in achieving a goal. Providing particular recommendation by using Social Network Services (SNS) with geographical location-aware feature is one of the interesting SIoT problems. This is mainly because the recommendation is usually constrained by several factors, which pose challenges to us on the available options, and hence having the right information that leads us in taking right decision at the right time is essential. The main contribution of this paper is providing a personalized recommendation framework which is suitable to be adopted in SIoT systems. We also analyzed the performances of the proposed framework with different datasets. Promising results are shown from the analysis.