Social Influence Maximization Problem Based On EPC Network

In recent years, by taking advantage of the rapid development of Internet of Things, EPC Network is being applied in more and more fields. EPC Network(Electronic Product Code Network), proposed by EPCglobal organization, is a kind of open network platform to share product information across company boundaries and business partners. Its basic information flow is captured from RFID tags on(or implanted on) products. Each of these RFID tags is given an EPC number which is a globally unique identifier. RFID readers are responsible for identifying tags and collecting information. EPC Network has now played an important role on anti-counterfeiting, tracing management and anti-out-zone-sale in many business fields. The research and development of EPC Network, however, do not value end user experience and their behavior data currently. In addition, EPC Network lacks an earning mode to reduce its operating costs.This thesis analyzes current domestic and foreign research status, and proposes a social influence maximization model for viral marketing based on EPC Network. This model aims to find the top influential people in a social network that maximize the spread of influence from learning query events in EPC Network performed by its end users. Thus, some manufacturer can develop a market for their products through the influence of these seed people. Our research work can be divided into the following main four steps: Learning user relationship from EPC query event, Integration with Social Network Site, Learning social influence probabilities from EPC query event and shop-perspective influence maximization. These four steps can be divided into two parts: the first three steps create the model while the last step is the analysis part of the model. In the first part, we first learn the physical user relationship from end users’ consuming behaviors and integrate with the virtual user relationship learnt from social network site. Then the influence probabilities are calculated from consuming behaviors to quantify the extent of influences between users.Firstly, in this paper the background and the conception of Internet of Things, EPC Network and viral marketing is introduced as well as the research contents and innovation points in this paper. Secondly, the current research status of EPC network and social influence maximization problem is analyzed along with their features, advantages and shortcomings. Thirdly, based on the data structure of EPC Network event and viral marketing requirements, a social influence maximization model aims to mine the top influential end users is proposed including the overall design of the model and each steps. Then the implementation of the model and its core parts is described. In the display unit of the model system, a smartphone APP is used to demonstrate the running effect. Finally, the summary of the full paper is made to show the unsolved problems and future works.