A Feature-Weighted Clustering approach for Context Discovery and Selection of Devices in IoT

Internet of Things (IoT) intended to connect various physical devices in multiple domains to offer high quality ondemand services. In this scope, identification of intended devices is remains a challenge because of heterogeneity and wide distribution. Context plays a significant role to enable provision of adequate services to the users based on their preferences. In addition, the context can help to adapt with the dynamic environment changes. Therefore, the aim of this paper is to address how the context can be discovered from IoT data, and its influence in recommending the IoT devices. For this, a weighted clustering mechanism is applied aiming to discover the informative contexts and recommended the devices to the user based on the context similarity. The proposed model is extensible, independent of domain and taking into account the constraints of the IoT like availability, applicability, etc. Further, this model is validated through a cross validation mechanism which shows accurate prediction of probable contexts.