| 摘要 |
A method and a system for analyzing moods of end-users of web-sites relative to the task graphs of the web-sites. Problems detected with end-user moods are fed back to business knowledgeable users of the web-site, to aid them in improving the web-site. In order to determine mood states of end-users, data is collected from various sources, which is sorted and filtered. Information is also obtained regarding end-user success on the various tasks that can be performed using the different pages of a web-site. Based on this information, the task model of the web-site may be augmented with new edges which reflect unexpected end-user path segments. The states that end-users can be in, include general states (e.g., shopping, purchased, or dead) and mood states (e.g., happy, confused, bored or angry). The factors used to determine end-users'states can include the time spent by the user on different web-pages, the links in a task graph followed by the user, how efficiently they progressed towards their objective (e.g. did they go in circles, or travel the shortest path between A and B) etc. The data obtained from the various sources may be weighted by several different factors such as the size of a web-page, and the native language of the end-user. The report generator generates reports in various formats regarding end-user states, and includes a linkage of problematic states of end-users with problems detected from a semantic analysis of the web site. |
