Bilişim Teknolojileri Dergisi, cilt.11, sa.2, ss.211-222, 2018 (Hakemli Dergi)
Nowadays, websites are used by billions of people. The identification of the visitor needs of website is very important for a person/a community/an organization or a brand to reach more people through a website, to be accepted by website followers, and so to achieve targeted success. These identified needs play a key role in improving a website in terms of design and content. The aim of this study is to provide a case study to identify visitor needs of a website by using in-site search and the apriori algorithm. In this context, a monthly web log file which is obtained from Kırklareli University website (www.klu.edu.tr) was used as data set. Analysis process is discussed in the context of CRISP-DM: CRoss-Industry Standard Process for Data Mining. It is observed that word/word groups of “Undergraduate Transfer”, “Re-enrollment”, “Syllabus”, “Course Registration”, “Tuition” and “Quota” mostly exist in searches performed by visitors in the results of analyzes especially during the month that the university registration process is done intensely. In the results of analyses performed with apriori algorithm, it is found that the searches of “Undergraduate Transfer” (1177 times), “Tuition” (889 times) and “Syllabus” (600 times) lead of all searches. Association rules such as “%60 of the visitors who searched Syllabus and Undergraduate Transfer, searched Tuition as well”, which allow the university website to serve better, have been shared in the study.