Abstract: |
The purpose of this study is to segment the Internet health information market of health counselling, disease consultation, health commodity shopping, and hospital information in Korea by using decision tree, a widely used data mining algorithm. Telephone survey with structured question was performed, and finally 8,656 completed interviews were used for the analysis out of 10,325 respondents of Kangwon province and Incheon city in Korea. The survey was conducted from July 2006 to October 2006. We used CHAID algorithm where chi-square statistics is used to find optimum split. Dependent variables are experience of the Internet health information access, disease consultation, health commodity shopping, and hospital selection, while independent variables are demographic data and health conditions. Among the 8,656 samples, 1,665 (19.2%) have used the Internet for health information search during the previous year. The main purposes of the Internet search was, allowing plural choice, for general health tips(64.2%), disease consultation(32.0%), health commodity shopping(23.7%), and hospital selection(19.3%). We found that each section of the Internet health information had its own devoted customers, and therefore customized market segmentation was strongly required. As a result, the Internet search pattern and customer behavior with health information in Korea was grasped, and the result would be useful to analysis the Internet health information market segmentation in Korea.
|