Fixing up links in social network based on Content Matching

Abstract

Social media has been the major source of information gathering. Analyzing social relationships is found to be a challenging task. In the present day scenario, the quality of analysis has significantly improved because of availability of large volume of data in any particular domain. Traditionally, the majority of information was earlier being produced by the few specialized sources or producers and consumed by a large number of customers.
However, the trend seems to be changed drastically over the time. But presently increasing number of platforms allow everyone to participate both in the production of information as well as consumption of information. The goal of this study is to efficiently distribute the information from producers to consumers. It is necessary to make higher relevance or maximize the overall relevance of the matched content from producer to the consumer while regulating the overall activities. So this problem serves as a matching problem. The two standard algorithms which are often used by other authors for making more relevance of product to the consumer are GreedyMR and StackMR. But proposed algorithm called b-Suitor works better than these two algorithms in terms of traversed edges and quality. The second goal is to predict the link between the nodes which does not exist in the current network but may present in the future network. For this link prediction, using the machine learning approach for that several supervised learning algorithms such as Support Vector Machine, Decision Tree, Naive Bayes, Bagging, Multilayer Perceptron, and K-Nearest Neighbor are applied and a comparative analysis has been carried out in order to find the best among them.