Please use this identifier to cite or link to this item:
標題: 提升圖書館推薦系統效率與視覺化的演算法
Algorithms for Improving the Efficiency and Visualization of Library Recommender System
作者: 張益龍
Chang, I-Lung
關鍵字: Personal Ontology
Ontology Visualization
OntoSphere 3D
OntoSphere 3D
出版社: 資訊科學與工程學系所
引用: [1] 邱宏彬、湯鎰聰、陳揮明(2005),「數位圖書館個人化檢索與推薦服務之設計與實作」,南華大學資訊研究學術期刊。 [2] 卜小蝶(1998)「淺析個人化服務技術的發展趨勢對圖書館的影響」。國立成功大學圖書館刊,2,63-73。 [3] Alessio Bosca, Dario Bonino, Marco Comerio, Simone Grega, Fulvio Corno :A reusable 3D visualization component for the semantic web. In: Proceedings of the twelfth international conference on 3D web technology (2007)89-96 [4] Akrivi Katifori, Constantin Halatsis, George Lepouras, Costas Vassilakis, Eugenia Giannopoulou: Ontology Visualization Methods—A Survey. In: ACM Computing Surveys (CSUR) Volume 39 , Issue 4 (2007). [5] 陳慧玲(2007),「植基於個人本體論的圖書館推薦系統—以中興大學圖書館為例」,中興大學資訊科學研究所,碩士論文。 [6]王煥宇(2009) ,「一個應用三角不等式的個人本體論相似度演算法」,中興大學資訊科學研究所,碩士論文。 [7]許正怡(2008) ,「植基於個人本體論模型與合作式過濾技術之中文圖書館推薦系統」,中興大學資訊科學研究所,碩士論文。 [8] Cristian Perez de Laborda, Stefan Conrad: Relational.OWL: a data and schema representation format based on OWL. In: Proceedings of the 2nd Asia-Pacific conference on Conceptual modelling - Volume 43, (2005) 89 – 96 [9] T. Gruber (1993). "A translation approach to portable ontology specifications". In: Knowledge Acquisition. 5: 199-199. [10] F. Arvidsson and A. Flycht-Eriksson. Ontologies I. Retrieved 26 Nov 2008. [11] I-En Liao, Shu-Chuan Liao, Kuo-Fong Kao, and Ine-Fei Harn, “A Personal Ontology Model for Library Recommendation System,” Proceedings of 9th International Conference on Asian Digital Library, S. Sugimoto et al. (Eds.), Lecture Notes in Computer Science, Vol. 4312, Springer-Verlag, November 2006, pp. 173-182. [12] Keith Andrews, Josef Wolte, Michael Pichler "Information PyramidsTM :A New Approach to Visualising Large Hierarchies" IEEE Visualization’97, Phoenix, Arizona, October 1997, pp. 49–52. 網路資源 [W1] OntoSphere 3D - User Guid,線上檢索日期:2010 年5 月 [W2] OWL Web本體語言概述 推薦標準,線上檢索日期:2010 年5 月 [W3] Ontology (information science) ,線上檢索日期:2010 年5 月 [W4] 本體論: 讓舉一反三變得可能,線上檢索日期:2010 年5 月 [W5] TouchGraph | Products: Navigator, 線上檢索日期:2010 年5 月 [W6] OntoViz線上檢索日期:2010 年5 月 [W7]IsaViz: A Visual Authoring Tool for RDF 線上檢索日期:2010 年5 月 [W8]SpaceTree 線上檢索日期:2010 年5 月 [W9]OntoTrack 線上檢索日期:2010 年5 月 [W10]Jambalaya 線上檢索日期:2010 年5 月 [W11]OntoSphere 線上檢索日期:2010 年5 月 [W12]Grokker線上檢索日期:2008 年7 月 [W13]MoireGraphs線上檢索日期:2010 年5 月 [W14]Touch-Graph 線上檢索日期:2010 年5 月 [W15]3D Hyperbolic Tree 線上檢索日期:2010 年5 月
摘要: 隨著資訊爆炸時代的來臨,個人化推薦服務成為各類資訊系統中,非常重要的元件。以中興大學為主的研究團隊,從2006年開始,針對圖書館的圖書借閱服務,建立了一個以個人本體論為基礎的圖書館推薦系統,並將之命名為PORE ( Personal Ontology REcommender System)。這個系統是目前所知,第一個將個人本體論應用在圖書推薦的系統,具有相當的前瞻性。然而,在實作上,這個系統仍有需要改善的地方。本論文便是針對其視覺化展現推薦結果,以及合作式過濾的效能,提出改善。 在推薦結果視覺化方面,本論文使用開放原始碼的工具OntoSphere 3D,將個人本體論以及其與推薦圖書的關係,以3D的方式呈現。在三度空間中,物件可以旋轉、收縮、移動、變化形狀,以更豐富的視覺效果,提供讀者更真實的查閱界面。在合作式過濾的執行效能方面,我們也透過預先分群的方式,將有重疊借閱分類紀錄的讀者先歸為一群,並在該群讀者中去找相似的讀者。在實驗中,新的演算法,將可以節省96%的運算次數。本研究的成果,使PORE系統的運作更為快速,並具有優良的讀者介面,對系統的推廣,將有極大的幫助。
In the era of information explosion, personalized recommendation services have become an important component for most information systems. Since 2006, the research team mainly consisting of researchers at National Chung-Hsing University (NCHU) proposed a recommender system called PORE (Personal Ontology REcommender System) for recommending library collections. The system is the first one in using personal ontology for library collection recommendation. However, there is still some weakness which should be improved in the PORE system. This thesis proposed methods on improving visualization of recommended results and the performance of collaborative filtering recommendation. For visualization of recommended results, this study developed a 3D interface using an open source tool, OntoSphere 3D, to show personal ontology, the relationship between personal ontology, and the recommended library collections. In the 3D model, users can rotate, shrink, or move recommended objects. As a result, the 3D interface provides a more user-friendly interface to the users of the PORE system. For the performance improvement of collaborative filtering recommendation, we proposed a method that clusters users with overlapping personal ontology and then finds similar users in the group. The experimental results show that the proposed method can save up to 96% of computation.
其他識別: U0005-3008201011362900
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.