New Model to evaluate User Reliability Dynamically for Social Network Services

摘要

With the rapid propagation of smart devices, the Social Networking Service (SNS) has become the most popular service in IT industry. In SNS, there is no classification between information producer and consumer, but all participants can produce new information or use information freely. According to a recent e-marketer's report, about 1.2 billion people in the world use one or more SNSs. However, with this dramatic growth, various adverse effects have been reported: for example, impersonation, unproven rumours, online personal attack, and exposure of privacy. A girl in America even killed herself because of personal attacks on an SNS. To prevent these adverse effects of the SNS, moral advancement of SNS users is naturally the best solution; however, this goal is practically very hard to achieve. Therefore, another solution is demanded, since the SNS has already become an essential service in our lives. For these reasons, this paper proposes a new model to dynamically evaluate user reliability for SNSs. The purpose of our new model is to quantify user reliability, and to provide it to all participants. Based on this quantification, an SNS user can easily determine who is good or bad, and then makes a good relationship or bans bad people. Consequently, our new model provides a criterion in the building of online relationship, and we expect that our new model may replace authentication procedure.