| 摘要 |
An invention for customizing a web site by gathering information from a web site visitor's sessions while the user is using the web site. The visitor's session information is compared with expected visitor behavior, based on the present structure of the web site. Using pre-programmed basic comparison rules and computer based mathematical models, matrices are used to represent statistical information about the visitor's sessions on the web site. The statistical information is used to extract visitor behavior which was unexpected (anomalies). Anomalies are grouped into recommendations. These recommendations are used to automatically customize the web site. In the alternative, information is provided to the web site administrator to customize the web site to be more efficient and visitor friendly, maximizing the operation of the Web site and promoting more frequent visits.
|
