Please use this identifier to cite or link to this item:
標題: 圖書館網站個人化推薦系統研究
A Personalized Recommendation System for Library Services
作者: 韓瑛馡
Harn, Ine-Fei
關鍵字: Library
Recommendation System
出版社: 資訊科學研究所
摘要: 摘要 如何提供讀者良好的服務,協助讀者獲取資訊、運用資訊,從而產生知識,是圖書館最重要的使命。而隨著網際網路的蓬勃發展,人們已漸漸習慣於網路上尋找自己想要的資訊,因此如何利用資訊科技及網路來提升圖書館的讀者服務相信是各圖書館努力的目標。 本研究以中興大學圖書館的館藏書目、讀者的借閱記錄及中國圖書分類法為基礎,以推薦系統中的內容導向過濾(Content-Based Filtering)的方法,搭配本體論(ontology)的推論,找出讀者真正有興趣的主題書目。當找出讀者興趣之後,將推薦主題清單依讀者興趣值高低做個人化的排序,於圖書館網站上推薦給讀者。再者,以建立各主題書目之語意索引的方式,協助使用者更快速的搜尋感興趣的相關主題。 本研究設計的個人化主題推薦系統,以站在個人本位及提供讀者主題知識的角度,運用Ontology與語意索引等技術配合圖書館特有的分類編目知識來提供使用者感興趣的主題館藏資訊。而實驗結果顯示,這套系統確實可以有效的依照讀者興趣將主題館藏推薦給讀者。
ABSTRACT The primary purpose of libraries is to provide desirable services for readers, to assist them in retrieving information and applying information, as well as to acquire knowledge. As the development of the Internet, people are gradually used to searching desired information online. It is believed that striving toward utilizing the information technology and the Internet to increase and broaden services for users is every library's challenge. The research takes the collections of the library and the readers' loan records of the National Chung Hsing University, as well as the New Classification Scheme for Chinese Libraries as the foundation. Then the research collocate the inference using ontology approach to find out the subject catalogue that readers are really interested. After that, the research can sort the obtained wish list for readers' reference by using the Content-Based Filtering Methodology of the Personalized Subject Recommendation System. In addition, the system establishes the semantic indexing for each subject collection so that users can search the relevant subjects promptly. The Personalized Subject Recommendation System of the research is proposed from the viewpoint of the individual and subject knowledge furnishing; and then the system applies Ontology and Semantic Indexing technologies as well as the cataloging & classification knowledge specialized for the library to provide collections that users are interested. The research results reveal that the system can effectively make recommendations to readers according to their interests.
Appears in Collections:資訊科學與工程學系所



Show full item record
TAIR Related Article

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.