Su Yuzhao,Niu Xiaotai,Zhao Yan
图书与情报. 2011, 31(05): 60-64.
Abstract (
)
Download PDF (
)
Knowledge map
Save
This paper analyses the processes and methods of web log data modeling through the research of personalized recommendation and data modeling theory, and points out that the normal web log can't meet the needs of personalized user-analysis, prediction and the recommended precision. It puts forward the process and method of customizing Web blog data modeling and sets up a prototype of customizing web log data model. We use data mining technique to analyze association rules, classifying and clustering. The experiment shows that the data quality of customized web log data can provide good quality to find out the association rules, the content classification and the user clustering, therefore, it can improve the precision of Personalized Recommendation. Moreover, it has its merits in simplifying data preprocessing and increasing multiple utilities.