Adding Temporal Dimension in Social Network by Using Link Analysis

Abstract

Social Media Systems (SMS) are included in the category of web-based systems. All the social media sites are merged by Social Media Systems. On account of having dynamic nature, these media sites like amazon flicker so on and so forth, are fronting interaction overhead predicament. In order to subjugate the predicament, the researchers have proposed the framework termed as Social Recommender System (SRS). The most pertinent information is filtered by these systems to target web user with the use of information filtering process and Folksonomy base structure. Folksonomy does not work alone, but with the amalgamation of the graph-based approach and content-based approach which support the older tags. The graph-based approach used for link analysis whereas content-based approach is used with the degree of relevance between the query and the document. The temporal dimension of users and tags are not deliberated by these approaches or techniques. As the user's interests modify with the passage of time, this work gives argument regarding the temporal dimension of the users and tags which encompass the high level of importance. ‘TimeFolkRank’ (TFR), a technique is also presented in this paper which uses link analysis. In addition, the temporal dimension of the users and tags are not deliberated by this technique. In this paper, the proposed model has been evaluated with FolkRank and Language models having the status of baseline models. On Bibsonomy Dataset, our work measure the recall, precision and F1 measure of the recommended tags and users. The proposed model’s results highlight the degree of importance over existing results through experimental results.