Web Usage Pattern Discovery: the Case of Addis Ababa University Official Web Site
No Thumbnail Available
Date
2012-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Web sites are one of the well known communication media where very huge information
could be disseminated for visitors‘ access. Since most of the organizations in the world
use web sites to communicate their business, smooth running and better service of the
web sites is fundamental point for the benefit of users and/or the organization.
In order to develop improved web site services, the extent to which users are satisfied on
the web site should be computed to evaluate the site. On the first case, the site designers
could bring amendments on the site by their expectation of users‘ desire. Secondly, users
of the site may be communicated through interview, questionnaire or observation to
gather the users‘ trend of the web site usage. Thirdly, researchers could undergo web
usage analysis on web server log files of the users‘ requests.
In this research, server log of Addis Ababa University web site was utilized to discover
web usage pattern using statistical analyzer and data mining applications. Access log file
was a target data source as it is expected to show users‘ usage trends.
The methodology proposed by Srivastava et al. (2000) was used in the research work
which comprised of raw log preparation, pattern discovery and pattern analysis. The well
known association mining algorithm, Apriori, was used to extract the common directories
that were accessed together. Meanwhile, PredictiveApriori was used in mining common
navigational sequences in the web site. Then, the discovered rules were analyzed in
consultation with the site map of the web site and the concerned site designers.
The research found that there are directories which are frequently visited in the web site.
Other findings are also discussed based on which recommendations are forwarded.
Description
Keywords
Web Site