Content Distribution Network Based On Data Mining Research And Applications

The whole name of CDN is "Content Delivery Network". Its aim is adding a new network architect to the internet, make the content of the network to the edge of the users, in order that the user can receive the content from the nearest, solve the problem of network crowded, increase the response speed of user visiting.Nowadays the most of China CDN design is based on experience modal and illegible design. It can't adapt to the highly changeable and complex network environment.This paper is base on CDN construct, with core of research data mining of file distribute network, design a new content delivery system, in order to solve the problem which makes users visit Web slowly, raise the efficiency of the whole system, easily manage the system.First mainly introduce use software to realize the system load balance. Include database server load balance, web server load balace, file server load balance. Use middle server and some new technology to realize it, support the base of cdn.Second detailed introduce the algorithm of data mining, and data warehouse which is built for solving delivery problem, the distill and application of metadata, use clustering mining algorithm, get userful data, in order to provide better according for delivery.At last introduce it's based on the data which is analysed by data mining, design a new delivery algorithm, main included the first aim and second aim of the delivery, the first delivery of file, one hour groups delivery, one day groups delivery, and one hour one group delivery. not only raise the efficiency of delivery, save the space of server, but also make the best effect for user play, enhance the function of system.