Research On Program Monitoring Of Video Websites Based On The Technology Of Flash Cookies

With the rapid development of internet,the requirement for WEB applications is higher and higher.And the application of Flash cookies is wider and wider.Because of its characteristics,flex with the program can be understood as a programming language.And it is compiled after the completion of common Flash format file.Breaking the static Webpage page,it accustoms to the pattern,and greatly enhances the interaction Webpage.But it also brings some problems in loss of click stream.Most web applications are used the web server log file analysis of user access patterns,such as abnormal monitoring.The advanced applications,including data mining,click stream time series analysis etc.The Flash is a download,update and UI,which does not require a web server communication.Of course,Flash cookies,which use HTTP protocol,can get some XML format data.But these data do not reflect the user preferences for the content.In order to solve this problem,the research gives us an efficient algorithm,which accurate collection of user for Flash of various operations.Studying on the existing multimedia retrieval gradually by the keywords retrieval,the way steers object retrieval methods,which names to video segments as input,fro m a mass of video libraries.Conventional retrieval method,which is based on visual words,ignore the relevance of video frames in time sequence.There is still space to improve the retrieval effectiveness.This paper,which is in the process of management of SVA Group log data,founds some video website program monitoring broadcast,monitoring rules,and an improved algorithm.Comparing with the K clustering algorithm,the new algorithm is the superiority,which is named as FLASH_BC algorithm.It can take into account the video frame sequence relationship,and maintain acceptable retrieval speed.In contrast,K means clustering algorithm in processing large data environment of video website program data of illegal program efficiency and speed of the flow exist defects recognition algorithm.