FP-Tree based Association Rule Mining in Academic Social Networks to Refine RDF Framework and FOAF

Abstract

In order to analyze large scale social networks, different strategies are being implemented. The traditional methods of data mining are getting transformed to be suitable to the requirements of the web based information available in different structured and unstructured formats. The process of web mining, a versatile methodology of data mining, involves modified mining techniques applied on documents spread across the World Wide Web. In this proposed work, the process of association rule mining is applied to Academic Social Networks, an offshoot of Social networks. This technique involves processing of profile pages of members stored in FOAF format and retrieving the association rules from the data to identify the strength of relationship amongst researchers. The RDF vocabulary, a standard format of web based identity representation is the building block of this entire framework. Traditional applications of association rule mining is referred in the context of Market Basket analysis, and its application is projected to analyze dense network and derive association patterns amongst members of the academic social network.