Nowadays, along with the high development of emerging computational paradigms, more and more populations have been involved into the social revolution across various intelligent systems, which results in dynamic user connections associated with a variety of social behaviors. The associated users with different properties, who can be regarded as one kind of information resources, have become increasingly important, especially in social knowledge creation and human intelligence utilization processes. In this study, we concentrate on user role identification based on their social connections and influential behaviors, in order to facilitate information sharing and propagation in social networking environments. Following the construction of a dynamic user networking model, we propose a networkaware method to identify four kinds of special users, who may play an important role in information delivery among a group of users, or knowledge sharing between pairs of users. A set of attributes and measures is proposed and calculated to identify and represent these users based on the analysis of their influence-related social behaviors and dynamic connections. Experiments and evaluations are conducted to demonstrate the practicability and usefulness of the proposed method using Twitter data. Analysis results show the effectiveness of our approach in identifying the distinct features of four kinds of users from the user networking model. Comparison experiments indicate that the proposed identification method outperforms two other related works. Finally, a questionnaire-based evaluation demonstrates the accuracy and efficiency of the proposed method in terms of finding these users in a real social networking environment.

Introduction

With the rapid development of emerging computing paradigms, such as Ubiquitous Computing, Social Computing, and Mobile Computing, we have been continuously experiencing a fast change from all walks of our work, life, learning and entertainment. People are joining together to publish personal messages, share individual experience, and exchange their own opinions through online social networking services. As more and more people have been engaged into this social networking environment, a large number of user generated contents, which contain a variety of human experience and social knowledge, have spread widely in a higher speed than ever before [1]. Comparing with the traditional information dissemination which fully depends on the popularity of posted contents, the information flow propagated cross social networks mainly relies on interactions among individuals and groups associated with various social relationships. That is, the dissemination speed, scale, and controllability are increasingly influenced by the highly connected users. In particular, individual users with different background knowledge (e.g., cognition, interest, reputation, and etc.), are playing a significant role not only in shaping public opinions, but also in expanding access to diversified personal contents more efficiently. Thus, it becomes a challenge but essential issue to dynamically identify types or roles of individuals, who may help to deliver human intelligence and socialized knowledge to the person in need via user relationship chains during information sharing, exchanging, and propagating processes.