Based Distance Education, Web Data Mining Technology

Distance education has been developed rapidly these days. While studying, students run through the Web pages. The information of these browsed pages is collected by Web servers automatically and saved in log files. Among the saved information are included students' interests, configuration of the site structure, etc. We can obtain useful and favorite clues by making data mining of these log files.In this paper, we start from with an example, from data collecting of log files, data preprocessing, forming correlative matrix, to normalizing. To classify similar learners into groups and connect to relative pages, fuzzy-similar-theory-based distance measure, so called Haming distance, is used to analyze the similarity between different students and different pages. There is also an analysis of the frequencies of access paths used by students to find certain path most frequently chosen. All of these results can be very helpful to direct the building of Web sites and improve the quality of distance education.This method is simple and easy to implement. It has low demands on data preprocess and no special requirement on the format of Web log files. And it is unnecessary to distinguish between users and sessions. It is one of the best methods to make data mining of the log files on educational Web sites, as it were.