Object Recommendation based Friendship Selection (ORFS) for navigating smarter social objects in SIoT

Abstract

Social Internet of Things (SIoT) paradigm contributes for the social relationship between the objects. SIoT autonomously handles diverse set of objects, performing similar responsibilities. Providing user friendly smarter objects for resource sharing is a prerequisite in any SIoT platform. An effective service discovery depends on multiple attributes of object centrality. In this work, an efficient Object Recommendation based Friendship Selection (ORFS) model for network navigability and social relationship management among smarter objects in SIoT is proposed. The proposed ORFS model empowers to create, communicate and manage trust based social objects. A Grey Wolf Algorithm based User Object Affiliation (GWA-UOA) mechanism for the Smarter Object Recommendation (SOR) is proposed. Then Object Friendship Selection (OFS) through network navigability using Maximum Ranked Neighborhood (MRN) approach is proposed. Finally, desired service is navigated through the established friendship link. The proposed social-driven relationship approach is validated using two distinct real-world datasets. Experimental results proved that the proposed ORFS model well performed for navigating smarter social objects in SIoT. ORFS model achieved an improved performance with various factors like MAE, RMSE, computational time, average path length, recall, precision and F1 score.