Cluster Based Detection and Analysis of Internet Topics

Internet topic detection and classification is an intelligent information access technology. It studies how to detect new events and classify sentiment of the content. Classical detection and analysis system of internet topics has low analysis efficiency and large process delay. The functions of cluster-based analysis system are internet data collection, real-time analysis and off-line data analysis. Experimental results show that the Average Job Time (AJT) and Average Waiting Time (AWT) for jobs in case of Service Cluster are comparatively lesser with respect to Physical Server, and the Service Cluster shortens the service failover time by 93.4%.