Public Opinion Intelligence Analysis Algorithm of Efficient Course Online Practice Platform based on Message Data Cleaning and Evolutionary Algorithm Mining

It is of great significance to effectively collect the emerging massive Internet information and discover the hotspot public opinion on the Internet. This paper studies the representation and characteristics of network public opinion, and proposes the idea of using network method, combining Hits algorithm and PageRank algorithm to mine public opinion hotspots. Carry out specific research on the collection sources, combine the existing technologies and models of domestic and foreign public opinion collection system research, and carry out systematic analysis and design according to the actual needs of the network public opinion monitoring platform of South China University; Data crawling, using multi-threaded parallel crawling to achieve real-time efficient crawling.