Contents Recommendation Method Using Social Network Analysis

With the recent tremendous increase in the volume of Web 3.0 content, content recommendation systems (CRS) have emerged as an important aspect of social network services and computing. Thus, several studies have been conducted to investigate content recommendation methods (CRM) for CRSs. However, traditional CRMs are limited in that they cannot be used in the Web 3.0 environment. In this paper, we propose a novel way to recommend high-quality web content using degree of centrality and term frequency-inverse document frequency (TF-IDF). In the proposed method, we analyze the TF-IDF and degree of centrality of collected RDF site summary and friend-of-a-friend data and then generate content recommendations based on these two analyzed values. Results from the implementation of the proposed system indicate that it provides more appropriate and reliable contents than traditional CRSs. The proposed system also reflects the importance of the role of content creators.