Link Classification and Tie Strength Ranking in Online Social Networks with Exogenous Interaction Networks

摘要

Online social networks (OSNs) have become the main medium for connecting people, sharing knowledge and information, and for communication. The social connections between people using these OSNs are formed as virtual links (e.g., friendship and following connections) that connect people. These links are the heart of today's OSNs as they facilitate all of the activities that the members of a social network can do. However, many of these networks suffer from noisy links, i.e., links that do not reflect a real relationship or links that have a low intensity, that change the structure of the network and prevent accurate analysis of these networks. Hence, a process for assessing and ranking the links in a social network is crucial in order to sustain a healthy and real network. Here, we define link assessment as the process of identifying noisy and non-noisy links in a network. In this paper (The work in this paper is based on and is an extension of our previous work.), we address the problem of link assessment and link ranking in social networks using external interaction networks. In addition to a friendship social network, additional exogenous interaction networks are utilized to make the assessment process more meaningful. We employed machine learning classifiers for assessing and ranking the links in the social network of interest using the data from exogenous interaction networks. The method was tested with two different datasets, each containing the social network of interest, with the ground truth, along with the exogenous interaction networks. The results show that it is possible to effectively assess the links of a social network using only the structure of a single network of the exogenous interaction networks, and also using the structure of the whole set of exogenous interaction networks. The experiments showed that some classifiers do better than others regarding both link classification and link ranking. The reasons behind that as well as our recommendation about which classifiers to use are presented.