建立层次结构用户兴趣模型的方法
|
摘要
为解决Internet上对于普通的用户日益严重的“信息迷失”和“信息过载”问题，智能信息检索和个性化信息服务成为当前信息服务研究的重点。作为其基础，一方面要研究如何应用人工智能的技术表示和组织Internet上的信息，另一方面要研究如何建立个性化用户模型，从而更清楚地获取和表示用户的兴趣和信息需求以及智能化地处理用户与检索系统的交互等。
本文研究建立层次结构用户兴趣模型的方法。层次结构的用户兴趣模型能够把用户的具体兴趣和笼统兴趣表示成一个连续的整体。从某种意义上说，具体的兴趣应该和短期的兴趣相对应，而笼统兴趣和长期的兴趣相对应，所以用户兴趣模型的这种表示方法为其应用带来了很强的灵活性。本文主要的工作和成果如下：
提出了一种层次结构用户兴趣模型的表示方法。一个层次结构用户兴趣模型是一个以单个兴趣向量为节点的树。
提出了一种建立层次结构用户兴趣模型的算法。用户兴趣模型是在用户浏览Web信息的过程中动态生成的，如何生成层次结构用户兴趣模型是算法的重点。
提出了一种对层次结构用户兴趣模型（即用户兴趣树）进行剪枝的方法。用户兴趣树中的节点不能无限制的增加，当节点数超出一定数目时就要对其进行剪枝。
建立层次结构用户兴趣模型的原型系统UMS实现。UMS以中文的Web页面为研究对象，它能监视用户的浏览过程，在对用户的浏览内容进行处理和分析的基础上建立用户兴趣模型。
As the solutions of the problems, such as “information overload” and “information labyrinth”, which are more and more serious to the ordinary users, intelligent information retrieval and personal information service have become the emphasis of the research of information service. Their foundations are to research how to apply the technologies of artificial intelligence to represent and organize the interest information, and how to build personalized user models, so that we can represent user's interest and information need clearly, and deal with the interaction between users and information retrieval systems intelligently.
This thesis researches the method of building hierarchical user interest model. The hierarchical user interest model can represent a continuum of specific to general interests of a user. In some sense, more general interests correspond to longer-term interests, while more specific interests correspond to shorter-term interests. So, this representation of user interest model provides flexibility for its application. The main contributions of this thesis are listed as below:
The representation of hierarchical user interest model. A hierarchical user interest model is a tree whose nodes are the single interest vectors.
An algorithm of building hierarchical user interest model. User modeling is a dynamic process during user browsing the web. How to generate hierarchical user interest model is the emphasis of the algorithm.
A method of pruning hierarchical user interest model (user interest tree). The nodes of user interest tree can’t increase without limit. When the number of the nodes exceeds given count, we must prune the tree.
The prototype system which is an implementation of hierarchical user interest modeling. It uses Chinese Web page as study object and can monitor the process of user browsing the web .On the base of processing and analyzing the content which
users are browsing It builds user interest model.
本文研究建立层次结构用户兴趣模型的方法。层次结构的用户兴趣模型能够把用户的具体兴趣和笼统兴趣表示成一个连续的整体。从某种意义上说，具体的兴趣应该和短期的兴趣相对应，而笼统兴趣和长期的兴趣相对应，所以用户兴趣模型的这种表示方法为其应用带来了很强的灵活性。本文主要的工作和成果如下：
提出了一种层次结构用户兴趣模型的表示方法。一个层次结构用户兴趣模型是一个以单个兴趣向量为节点的树。
提出了一种建立层次结构用户兴趣模型的算法。用户兴趣模型是在用户浏览Web信息的过程中动态生成的，如何生成层次结构用户兴趣模型是算法的重点。
提出了一种对层次结构用户兴趣模型（即用户兴趣树）进行剪枝的方法。用户兴趣树中的节点不能无限制的增加，当节点数超出一定数目时就要对其进行剪枝。
建立层次结构用户兴趣模型的原型系统UMS实现。UMS以中文的Web页面为研究对象，它能监视用户的浏览过程，在对用户的浏览内容进行处理和分析的基础上建立用户兴趣模型。
As the solutions of the problems, such as “information overload” and “information labyrinth”, which are more and more serious to the ordinary users, intelligent information retrieval and personal information service have become the emphasis of the research of information service. Their foundations are to research how to apply the technologies of artificial intelligence to represent and organize the interest information, and how to build personalized user models, so that we can represent user's interest and information need clearly, and deal with the interaction between users and information retrieval systems intelligently.
This thesis researches the method of building hierarchical user interest model. The hierarchical user interest model can represent a continuum of specific to general interests of a user. In some sense, more general interests correspond to longer-term interests, while more specific interests correspond to shorter-term interests. So, this representation of user interest model provides flexibility for its application. The main contributions of this thesis are listed as below:
The representation of hierarchical user interest model. A hierarchical user interest model is a tree whose nodes are the single interest vectors.
An algorithm of building hierarchical user interest model. User modeling is a dynamic process during user browsing the web. How to generate hierarchical user interest model is the emphasis of the algorithm.
A method of pruning hierarchical user interest model (user interest tree). The nodes of user interest tree can’t increase without limit. When the number of the nodes exceeds given count, we must prune the tree.
The prototype system which is an implementation of hierarchical user interest modeling. It uses Chinese Web page as study object and can monitor the process of user browsing the web .On the base of processing and analyzing the content which
users are browsing It builds user interest model.
引文
[1] T.Joachims, D.Freitag, and T.Mitchell.WebWatcher: A Tour Guide for the World Wide Web. In Proc.IJCAI'97,August1997.
[2] E.Casasola.ProFusion PersonalAssistant:an agent for perso-
nalized information filtering on the WWW.Master's thesis,The University of Kansas,Lawrence,KS,1998.
[3] H.Lieberman.Letizia: An agent that assists Web browsing. In Proc.Intl.Conf.on AI，Montreal，Canada， August 1995.
[4] 潘金贵，胡学联，李俊，张灵玲.一个个性化的信息搜索Agent的设计与实现.软件学报，2001，Vol.12.No.7：1074-1079.
[5] 李俊，张灵玲，周文辉，潘金贵.一个智能用户接口Agent设计与实现.软件学报，1999，Vol.10，No.8：835-842.
[6] 汪晓岩，胡庆生，李斌，庄镇泉.面向Internet的个性化智能信息检索.计算机研究与发展，1999年第36卷第9期：1039-1046.
[7] Lewis, D. D. “An evaluation of phrasal and clustered representations of a text categorization task”, In Proceedings of SIGIR-92, 15th ACM International Conference on Research and Development in Information Retrieval, Kobenhaven, DK, 1992： 246-254.
[8] Kobsa, A. “User Modeling: Recent work, prospects and Hazards”, In Adaptive User Interfaces: Principles and Practice Schneider-Hufschmidt, M. Kühme, T. Malinowski, U. (ed) North-
Holland 1993.
[9] Keogh, E. J. Pazzani, M. J. “Relevance Feedback Retrieval of Time Series Data”, In Proceedings of SIGIR '99, Aug 1999, Berkley, CA USA.
[10] Agrawal, R. Srikant, R. “Fast Algorithms for Mining Association Rules”, Proceedings of the 20th International Conference on Very Large Data Bases, VLDB, 1994.
[11] 范明，孟小峰等译.数据挖掘概念与技术.机械工业出版社，2001年8月第1版:223-259.
[12] van Rijsbergen, C. J. “Information Retrieval (Second Edition)”, Butterworths, 1979.
[13] Mitchell, T. M. “Machine Learning”, McGraw-Hill, 1997.
[14] http://www.ecs.soton.ac.uk/~sem99r/thesis-final.pdf
[15] 李晓黎，刘继敏，史忠植.基于支持向量机与无监督聚类相结合的中文网页分类器.计算机学报,第21卷，第1期：2001.
[16] 《汉语信息处理词汇01部分；基本术语（GB12200.1-90）》，中国标准出版社，1991年.
[17] 朱德熙.《语法讲义》，商务印书馆，1982年
[18] 殷建平.汉语自动分词方法[J]. 计算机工程与科学，1998，第3 期：60-66.
[19] Cooper W S.Getting beyond Boole.Information Processing & Management.1988 Vol.24 (3).
[20] Salton G and Lesk M E.Computer evaluation of Index and text processint. Association for computing Machinery.1968 Vol.15 (1).
[21] 罗可等.数据挖掘中聚类的研究[J]. 计算机工程与应用，2003 ，第20期:182-184.
[22] 朱 明.数据挖掘.中国科学技术大学出版社，2002年5月第一版：129-138.
[23] Liren Chen,Katia Sycara.WebMate:A Personal agent for Browsing and Searching.proceedings of the second international conference on Autonomous Agent,1998.
[24] QIAN Wei-ning,ZHOU Ao-ying.Analyzing Popular Clustering Algorithms from Different Viewpoints.软件学报，Vol.13,No.8:1382-1394.
[25] 卢超，梅卫峰，等.基于感兴趣度的WWW个性化信息发现系统.计算机科学，2002，Vol.29，No.10:67-69.
[26] 欧洁，林守勋，刘桂林.个性化智能信息提取中的用户兴趣发现.计算机科学，2001，Vol.28，No.3:112-115.
[27] Cher L, Sycara K.WebMate: A Personal Agent for Browsing and Searching. Http//www.cs.cmu.edu.
[28] Byoung-Tak Zhang, Young-Woo Seo. Personalized Web-Document Filtering Using Reinforcement Learning. http://citeseer. nj.
nec.com/correct/454621.
[29] Young-Woo Seo, Byoung-Tak Zhang.A Reinforcement Learning Agent for Personalized Information Filtering. 2000, Proc. of Int'l Conf. on Intelligent User Interface (IUI '2000).
[30] L. Ardissono, L. Console and I. Torre. An adaptive system for the personalized access to news. AI Communications, vol. 14, No.3:129-147..IOS Press, Amsterdam, 2001.
[31] Zamir, O. & Etzioni, O. Web document clustering: a feasi-
bility demonstration.1998,Proceedings of the ACM SIGIR Confer-
ence on Research and Development in Information
Retrieval:46-54.
[32] Kim, H. R. and Chan, P. K.. Learning implicit user interest hierarchy for context in personalization. In Proceedings of Intelligent User Interfaces (IUI'03), pages 101-108, Miami, Florida, USA. ACM.
[2] E.Casasola.ProFusion PersonalAssistant:an agent for perso-
nalized information filtering on the WWW.Master's thesis,The University of Kansas,Lawrence,KS,1998.
[3] H.Lieberman.Letizia: An agent that assists Web browsing. In Proc.Intl.Conf.on AI，Montreal，Canada， August 1995.
[4] 潘金贵，胡学联，李俊，张灵玲.一个个性化的信息搜索Agent的设计与实现.软件学报，2001，Vol.12.No.7：1074-1079.
[5] 李俊，张灵玲，周文辉，潘金贵.一个智能用户接口Agent设计与实现.软件学报，1999，Vol.10，No.8：835-842.
[6] 汪晓岩，胡庆生，李斌，庄镇泉.面向Internet的个性化智能信息检索.计算机研究与发展，1999年第36卷第9期：1039-1046.
[7] Lewis, D. D. “An evaluation of phrasal and clustered representations of a text categorization task”, In Proceedings of SIGIR-92, 15th ACM International Conference on Research and Development in Information Retrieval, Kobenhaven, DK, 1992： 246-254.
[8] Kobsa, A. “User Modeling: Recent work, prospects and Hazards”, In Adaptive User Interfaces: Principles and Practice Schneider-Hufschmidt, M. Kühme, T. Malinowski, U. (ed) North-
Holland 1993.
[9] Keogh, E. J. Pazzani, M. J. “Relevance Feedback Retrieval of Time Series Data”, In Proceedings of SIGIR '99, Aug 1999, Berkley, CA USA.
[10] Agrawal, R. Srikant, R. “Fast Algorithms for Mining Association Rules”, Proceedings of the 20th International Conference on Very Large Data Bases, VLDB, 1994.
[11] 范明，孟小峰等译.数据挖掘概念与技术.机械工业出版社，2001年8月第1版:223-259.
[12] van Rijsbergen, C. J. “Information Retrieval (Second Edition)”, Butterworths, 1979.
[13] Mitchell, T. M. “Machine Learning”, McGraw-Hill, 1997.
[14] http://www.ecs.soton.ac.uk/~sem99r/thesis-final.pdf
[15] 李晓黎，刘继敏，史忠植.基于支持向量机与无监督聚类相结合的中文网页分类器.计算机学报,第21卷，第1期：2001.
[16] 《汉语信息处理词汇01部分；基本术语（GB12200.1-90）》，中国标准出版社，1991年.
[17] 朱德熙.《语法讲义》，商务印书馆，1982年
[18] 殷建平.汉语自动分词方法[J]. 计算机工程与科学，1998，第3 期：60-66.
[19] Cooper W S.Getting beyond Boole.Information Processing & Management.1988 Vol.24 (3).
[20] Salton G and Lesk M E.Computer evaluation of Index and text processint. Association for computing Machinery.1968 Vol.15 (1).
[21] 罗可等.数据挖掘中聚类的研究[J]. 计算机工程与应用，2003 ，第20期:182-184.
[22] 朱 明.数据挖掘.中国科学技术大学出版社，2002年5月第一版：129-138.
[23] Liren Chen,Katia Sycara.WebMate:A Personal agent for Browsing and Searching.proceedings of the second international conference on Autonomous Agent,1998.
[24] QIAN Wei-ning,ZHOU Ao-ying.Analyzing Popular Clustering Algorithms from Different Viewpoints.软件学报，Vol.13,No.8:1382-1394.
[25] 卢超，梅卫峰，等.基于感兴趣度的WWW个性化信息发现系统.计算机科学，2002，Vol.29，No.10:67-69.
[26] 欧洁，林守勋，刘桂林.个性化智能信息提取中的用户兴趣发现.计算机科学，2001，Vol.28，No.3:112-115.
[27] Cher L, Sycara K.WebMate: A Personal Agent for Browsing and Searching. Http//www.cs.cmu.edu.
[28] Byoung-Tak Zhang, Young-Woo Seo. Personalized Web-Document Filtering Using Reinforcement Learning. http://citeseer. nj.
nec.com/correct/454621.
[29] Young-Woo Seo, Byoung-Tak Zhang.A Reinforcement Learning Agent for Personalized Information Filtering. 2000, Proc. of Int'l Conf. on Intelligent User Interface (IUI '2000).
[30] L. Ardissono, L. Console and I. Torre. An adaptive system for the personalized access to news. AI Communications, vol. 14, No.3:129-147..IOS Press, Amsterdam, 2001.
[31] Zamir, O. & Etzioni, O. Web document clustering: a feasi-
bility demonstration.1998,Proceedings of the ACM SIGIR Confer-
ence on Research and Development in Information
Retrieval:46-54.
[32] Kim, H. R. and Chan, P. K.. Learning implicit user interest hierarchy for context in personalization. In Proceedings of Intelligent User Interfaces (IUI'03), pages 101-108, Miami, Florida, USA. ACM.