Class-based request control system in multimedia contents service

This paper proposed the system and the algorithm for controlling clients' service requests using the CRM of multimedia contents service through client segmentation, based on a site's revenues. Thus, clients with higher profitability will have more chances of accessing media server resources. Likewise, controlling the service requests of all clients will enable controlling the use of network resources. To realize the system. K-Means Method was applied to the analysis of the clients' data as the first step. The clients were classified into four classes (First Class, Second Class, Third Class and Fourth Class). Two experiments were conducted to evaluate the performance of the system. The first experiment is the comparison of the number of admitted streams of proposed algorithm with other algorithm. In the second experiment, the renewal rates in the two cases were compared. Specifically, the renewal rates of the group of clients with expired contracts where the new algorithm was applied and those of the control group were evaluated. The renewal rate was higher in the group where the new algorithm was applied, suggesting an improvement in the satisfaction level of clients and revenues.