Research And Application Of Web Mining In Personalized Service Of Network

With the rapid development of information technology,internet-based applications are increasingly being used on the internet and soon accumulates a great deal,and different types of content-rich information.At present the network has become the primarily way that people access information and services.However,because of the features that it has many different types of network information,updating too fast,and too many useless spam,so that results in internet users are difficult to quickly obtain useful information through the Web page that they want.How can efficient and reliable to meet the information needs of customers,to provide better information services,more and more people's pay attention on it.Previously internet information services mainly through Web search engines retrieved information,then filtered information to the user on demand.But this method did not take into account the user on each person's background,habits,and the different visiting purpose.They provided a single popular information service that was difficult to meet the demand,so they were needed to provide targeted,personalized information services.To solve this problem,this thesis researched on how to use Web data mining technology in the network personalized service.This thesis introduces the Web data mining theory and application scope,and focuses on research and analysis of Web data mining,including data resources,mining processes and key mining technologies.Secondly,the thesis deeply researches on how will the association rules algorithm applied to the network to better personalized service,analysises the the shortcomings of Apriori algorithm,and proposes the establishment of a temporary table based on transaction database.And in accordance with the following after each generate frequent item sets a number of items required to dynamically update the temporary table to dramatically reduce the scanning workload and improve the efficiency of ideas.Quick_Apriori algorithms devised to be applied to mining network users to access network resources habits,and to find out is to access the resource association rules between them,resulting in strong association rules model.And examples of Quick_Apriori algorithm implements with the original algorithm are compared and analyzed,summarizes the advantages and disadvantages of the improved algorithm.Then,the thesis introduces thecontent of the network personalized service,points out the shortcomings still exist.A new optimization method of personalized service based on Web data mining technology is put forward,including the module design of Web data mining,design of personalized service platform system framework based on Web data mining and design of personalized service process.Finally,using the Quick_Apriori algorithm as the main algorithm of true user data in a university teaching resources website platform for the mining,the results show that this algorithm can mining association rules from the point of view to better achieve the personalized service.