Optimization-assisted personalized event recommendation for event-based social networks

Abstract

The usage of Online Social Networks (OSN) is promptly increasing in recent days. Particularly, Social Media networks permit individuals to share, communicate, observe and comment on diverse kinds of multimedia content. These phenomena produce a massive quantity of data that shows Big Data features, primarily owing to their large volume higher change rate, and inherent heterogeneity. In this viewpoint, Recommender Systems are established for helping users to discover “what they need within this ocean of information”. Here, this paper intends to design a novel personalized event recommendation approach, which deploys the “multi-criteria decision making (MCDM) approach” for ranking the events. In the adopted model, the preference schemes are built to compute categorical, geographical, temporal and social influences. Moreover, a personalized weight is approximated for every criterion (i.e., all influences). However, the major work deals with the estimation of personalized weight, and for this, new automated weight estimation is introduced via Weight oriented Grey Wolf Optimization (W-GWO) algorithm. Thereby, the dominance intensity measures is computed by exploiting the personalized criterion's weight of every criterion and the alternatives are given ranks depending on approximated dominance intensity measures for recommending the top-ranked events. Eventually, the supremacy of the adopted method is validated over other existing approaches in terms of various measures.