인구통계학적 정보와 연관성규칙을 이용한 협동적 추천시스템의 개발 : 백화점 데이터를 중심으로 = Development of a collaborative recommendation system using demographic data and association rule

다국어 초록 (Multilingual Abstract)

Among overflowing information because of the advancement of Internet, We want to get the most suitable one easily, quickly, precisely. Just like the case of E-commerce, customers would like to find the right items for them and sellers hope to find out profitable customers on their goods. It has been increasing the importance of automatic search systems, recommendation systems, and many studies are now under way.
There is a lot of distinction between each domain therefore, it would be different for customers to purchase items to their taste. In this study, we developed recommendation system using CBF, CF, Association Rules, K-Means clustering algorithms given the characteristics of the items of the department store.
The result indicated that the sample, which was done in connection with association rules, showed the best outcome than any others.

목차 (Table of Contents)

• 목차 = i
• 1장 서론 = 1
• 1.1 연구 배경 = 1
• 1.2 연구 목적 = 2
• 1.3 연구의 구성 = 6
• 2장 선행 연구 = 7
• 2.1 협력적 추천(Collaborative Filtering) 추천기법 = 8
• 2.2 내용-기반 추천(Content-based filtering) 기법 = 11
• 2.3 인구통계학적 추천(Demographic Filtering) 기법 = 12
• 2.4 연관성 규칙(Association Rule) = 13
• 2.5 K-평균 군집화(K-Means Clustering) = 14
• 2.6 브랜드의 의미와 기능 = 16
• 3장 인구통계적 정보를 이용한 추천시스템 = 20
• 3.1 선호 카테고리 추출 = 22
• 3.2 협동적 추천시스템 = 26
• 3.3 연관성 규칙의 적용 = 30
• 4장 실험 및 실험 결과 = 31
• 4.1 실험 자료 = 31
• 4.2 평가 방법 = 32
• 4.3 실험 결과 = 33
• 5장 결론 및 추후 연구 = 40
• 참고문헌 = 42
• ABSTRACT = 45