The Construction of a Recommendation System for Personalized Knowledge Web Pages

• Main Authors: Shen-Hung Liu, 劉信宏
• Other Authors: Ho-Chuan Huang
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/92598150391825234070

碩士 === 國立高雄應用科技大學 === 資訊管理系 === 99 === In recent years, the development of recommendation systems has become an important topic in research areas. However, there are still some unsolved issues in recommendation systems, such as articles recommendation based only on keywords in the article, while ignoring the semantic of the article, which will decrease the accuracy of recommendation. Therefore, the main purpose of this study is to build a personalized recommendation system for knowledge Web pages recommendation. Both domain ontology and personal ontology are used to express domain knowledge and user's preference, individually. In addition, data mining technique is also used to discover the relationship among concepts extracted from the personal ontology for the support of knowledge Webpage recommendation. The proposed system predicts the users' preferences by analyzing the past browsing records and ratings of individual users. After acquiring users' preferences, the personal ontologies are constructed and are used to discover the relationship among concepts by using the association rule algorithm. Finally, the system provides the personalized knowledge web pages to users based on association rules. The results showed that the proposed system provides the more precise result of Webpage recommendations than the traditional collaborative filtering recommendation systems.