Video measurement approach with classification based on service performance clustering

The Information Technology is improving these days, which pushes users' experience for video services to improve rapidly. In this trend, to improve the quality of experience (QoE) of online video service, a Measuring and Recommending System for Online Video Service (MCS) have been developed. In some traditional method, researchers classify resources with their video with the video attribute to reduce the performance measurement scale. Therefore, we proposed a video measurement approach based on the cluster analysis. In this approach, videos are clustered by their service performance. Just limited videos, which are selected from each cluster, would be measured instead of measuring all of them. A contribution in this work is that the accuracy shall be maintained significantly in this approach when compared with using the tradition way. In this experiment, the accuracy maintained when 70 percent costs had been reduced.