Exploring Users Navigational Behavior Using Web Usage Mining: The Case Of Ethiopia Commodity Exchange Official Website
No Thumbnail Available
Date
2015-06-04
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
The use of websites for distributing information is one of the most common media in today’s global market. It changes the business environment and has dramatic impact to build and maintain customer relationships through online activities to facilitate the exchange of ideas, products, and services that satisfy the goals of the organization. ECX is a business organization that works to revolutionize Ethiopia’s tradition bound agriculture to the global market. It has an extensive web site which provides users with access to services, products, and to advertise their organization. Despite the website’s potential of providing the information to the visitor, however, the website of ECX does not guarantee to being delivered and understood correctly. As a result, exploring user navigation behavior is expected to redesign the website based on user requirement and experience.
Web Usage Mining is the process of applying statistical analysis and data mining techniques to discover interesting usage navigation patterns of website. To explore usage patterns of the ECX official website the researcher followed Web usage mining process such as data collection, data preprocessing, pattern discovery and pattern analysis. The web server access log prepared by using log file viewer tool to clean irrelevant record from the log data, user and session identified by using Web log storming tool, preprocessed log record converted into the form appropriate for pattern discovery tool by using MYSQL statements.
After preprocessing of log file experiments conducted using statistical analysis with web log storming and association rule mining with Apriori and FPGrowth algorithm. The result of statistical analysis shows that half of the user of ECX website starts navigation at the root page (www.ecx.com.et/), others are directly access the page they want to visit with the help of search engine. More than two third (2/3) of visitors of ECX website exit from entry page without visiting the other page, others navigate other page before exiting from the entry page. Some visitors visit the same page or link frequently within the same session. Country wise visitor’s analysis shows that Iceland, Ethiopia, United States, China, and United Kingdom are the most frequent visitor countries. Most frequent visitor countries looks for coffee, white pea, maize and seasame. The most frequent page that case to error response is root page, which display partial content of the page (error type 206). The path analysis from statistical analysis and association rule mining shows that most of the pages of ECX website are not accessed together accept root and home page of the website. The major challenges that involved in this study are preprocessing of log file due to its large, noisy, and complex nature of log record, and identifying rules and patterns that are potentially interesting. Finally recommendation were done for decision makers, web designers, and further researchers to improve the website.
Description
Keywords
Exchange Official Website ;Behavior Using Web Usage Mining