전자상거래에서 상품 추천을 위한 웹 개인화 방안에 관한 연구 = A Study on Web Personalization for the Recommendation of Commodities in Electronic Commerce

다국어 초록 (Multilingual Abstract)

Numerous sites are in business on the internet and the number of sites are increasing at a higher rate. In order to acquire competitive advantage under this environment, we need to adopt effective and differentiative marketing strategy. One to one marketing is considered one of the possible alternatives. It is very crucial to collect information about the customer activities on the internet and to manage the database obtained through these data collection activities.
Click stream data shows the customer activity paths on the internet. It contains the information about what paths they followed to reach a specific site, what kind of pages they browsed through, what advertisement icons they clicked, and which customer eventually ended up with actual purchases. Therefore, click stream data is a very fertile source of information that can be used to the promotion of marketing, customer loyalty, and can be utilized to increase the rate of actual purchase.
In this paper, we applied data mining techniques on the click stream data to discover association rules. A clustering algorithm has been deployed to develop strategies as to how to personalize each web pages that will lead to the most efficient marketing activities, ensuring higher customer retention rate, higher customer loyalty, and eventually leading to higher ratios of actual purchases to visit.