사용자 행동 패턴 분석을 이용한 규칙 기반의 컨텐츠 사이트 관리 모델
- 원문 URL
Recently, digital content sites adopted pay contents service as a business model for profitability elevation. Due to characteristics of digital contents, such as illegal contents duplication or ID fraud are happened. Existing content protection methods were mainly based on two approaches: detecting illegal intrusion or unauthorized use through network or database, protecting content itself. In this thesis, we suggested a model for detecting a user with an abnormal behavior based on user access pattern to protect a digital contents in pay content site. For this model, syntactic and semantic detection rules are defined to analyze behavior patterns from access logs. And abnormal users are presumed by user behavior pattern deviated from normal range of detection rules. To prove efficiency of proposal model, we analysed user access log from a online recruit site with 12,682 users. As result of experiment, 390 users showed abnormal or singularity pattern. They showed singular behavior to access discordant contents with user's profile or to decrease number of visit time after subscription. The proposed model can be used in application of retention strategy in eCRM System by predicting churn customers. We proposed a new method for digital content management with user behavior pattern analysis. Using our model, we can manage users efficiently by monitoring concentrically singular user group and develop the system that manage contents and users in pay content sites.