多协同即时信息检索及其在电子教室中的应用
|
摘要
随着信息技术的快速发展和信息资源的极大丰富,工作中充分利用已有后备信息正成为提高工作和学习效率的一个重要途径和目标。即时信息检索正是利用后台的信息资源主动为用户提供服务的一种新的研究工作,通过软件代理感知系统的环境与当前的任务,感知的信息经过后台的数据库或者搜索引擎进行信息检索以后,返回相关信息参考条目。它是一种信息检索主动辅助任务的应用模式。
本文的工作是多协同即时信息检索,集中在如下两个方面:
第一,提出了多协同即时信息检索体系结构和设计框架。首次将数据库和知识库的双库协同,多代理协同,多向量协同同时用于即时信息检索。在双库协同系统中增加了缓冲知识库以提高运行速度,在多代理系统中考虑了代理通信,用KOML语言对通信原语进行了定义,检索中充分利用了兴趣向量和结果向量,并分析了这种设计框架。结论是对查准率有更多的优化,系统的实用面有很大的扩大。
第二,提出了一种多协同即时信息检索的应用模式。将多协同即时信息检索用于学习领域,探讨了在电子教室中的应用。对于载体为文本的任务,如阅读活动或者书写活动,提出多协同即时信息检索的电子教室并进行了体系结构的设计,给出了从上下文中得到关键词的算法,并且领域排序算法进行了设计,从内容层面对基于模板的匹配算法进行了设计,用(关键词,相关词,词汇表)分权重匹配的方法,对课程模板、兴趣模板进行了设计,根据匹配值的大小排序。
最后,给出了结论,多协同即时信息检索推进了Bradley Rhodes的工作,增加了信息感知的来源,提高了即时信息检索的效率和效果,将Agent从概念上进行了扩充,系统可以使用在多用户系统中,多协同即时信息检索电子教室对于改进学生的自主学习具有重要意义。
未来的工作,多协同即时信息检索电子教室的性能还需改进,基于性能来改进设计,提高电子教室的即时信息检索性能。在感知方式上,将文本感知扩大到媒体感知、综合感知。在应用上,还可向数字图书馆,数字博物馆扩展。除了用户兴趣之外,其它反映用户要求的心理内容在即时信息检索中的应用也值得进一步研究。
With the rapid development of information technology and more and more rich information resource, it has become an object and method to make great use of information existed. In this kind of application, there is Just-In-Time Information Retrieval (JITIR), a new theory. JITIR by software agent senses the environment and immediate task that the user is doing. The information sensed is retrieved by back-end database or search engine. And back-end returns to useful relevant items for user as reference. It is a kind of computer application pattern that information retrieval aids task.
Multi-cooperation Just-In-Time Information Retrieval (MCJITIR) is proposed. The cooperation is cooperation of DM and KDD, cooperation of agents and cooperation of vectors. In cooperation of DM and KDD, to increase response speed, the buffer knowledge database is added to the cooperation. In multi-agent, Communication of agent was paid attention to. The design framework is analyzed. The interesting vector and outcome vector are used. It is concluded that precision can be improved and the system can be used more widely.
One application pattern of MCJITIR is presented. MCJITIR is used in learning field. Application of MCJITIR in E-classroom is explored. For task based on text like reading and writing, the architecture design is given. The algorithm acquired keywords from context was designed. Ranking algorithm in fields based on template was presented. In the template, keyword, relevant words and vocabulary are weighted differently. According to the template, the match and rank is finished.
Last, the conclusion is given. MCJITIR advances work of Bradley Rhodes. MCJITIR adds the more sensing resource and improves the efficiency and effect. The conception of agent is expanded. MCJITIR can be used in multi-user system. The E-classroom using MCJITIR is of great meaning to enhance initiative learning.
In future work, the design need be improved based on evaluation of performance. In the method, the media sense deserves research. Sense of text and media may be integrated. To application fields, digital library and digital museum can use MCJITIR. The other mental contents in MCJITIR need be studied and explored further.
本文的工作是多协同即时信息检索,集中在如下两个方面:
第一,提出了多协同即时信息检索体系结构和设计框架。首次将数据库和知识库的双库协同,多代理协同,多向量协同同时用于即时信息检索。在双库协同系统中增加了缓冲知识库以提高运行速度,在多代理系统中考虑了代理通信,用KOML语言对通信原语进行了定义,检索中充分利用了兴趣向量和结果向量,并分析了这种设计框架。结论是对查准率有更多的优化,系统的实用面有很大的扩大。
第二,提出了一种多协同即时信息检索的应用模式。将多协同即时信息检索用于学习领域,探讨了在电子教室中的应用。对于载体为文本的任务,如阅读活动或者书写活动,提出多协同即时信息检索的电子教室并进行了体系结构的设计,给出了从上下文中得到关键词的算法,并且领域排序算法进行了设计,从内容层面对基于模板的匹配算法进行了设计,用(关键词,相关词,词汇表)分权重匹配的方法,对课程模板、兴趣模板进行了设计,根据匹配值的大小排序。
最后,给出了结论,多协同即时信息检索推进了Bradley Rhodes的工作,增加了信息感知的来源,提高了即时信息检索的效率和效果,将Agent从概念上进行了扩充,系统可以使用在多用户系统中,多协同即时信息检索电子教室对于改进学生的自主学习具有重要意义。
未来的工作,多协同即时信息检索电子教室的性能还需改进,基于性能来改进设计,提高电子教室的即时信息检索性能。在感知方式上,将文本感知扩大到媒体感知、综合感知。在应用上,还可向数字图书馆,数字博物馆扩展。除了用户兴趣之外,其它反映用户要求的心理内容在即时信息检索中的应用也值得进一步研究。
With the rapid development of information technology and more and more rich information resource, it has become an object and method to make great use of information existed. In this kind of application, there is Just-In-Time Information Retrieval (JITIR), a new theory. JITIR by software agent senses the environment and immediate task that the user is doing. The information sensed is retrieved by back-end database or search engine. And back-end returns to useful relevant items for user as reference. It is a kind of computer application pattern that information retrieval aids task.
Multi-cooperation Just-In-Time Information Retrieval (MCJITIR) is proposed. The cooperation is cooperation of DM and KDD, cooperation of agents and cooperation of vectors. In cooperation of DM and KDD, to increase response speed, the buffer knowledge database is added to the cooperation. In multi-agent, Communication of agent was paid attention to. The design framework is analyzed. The interesting vector and outcome vector are used. It is concluded that precision can be improved and the system can be used more widely.
One application pattern of MCJITIR is presented. MCJITIR is used in learning field. Application of MCJITIR in E-classroom is explored. For task based on text like reading and writing, the architecture design is given. The algorithm acquired keywords from context was designed. Ranking algorithm in fields based on template was presented. In the template, keyword, relevant words and vocabulary are weighted differently. According to the template, the match and rank is finished.
Last, the conclusion is given. MCJITIR advances work of Bradley Rhodes. MCJITIR adds the more sensing resource and improves the efficiency and effect. The conception of agent is expanded. MCJITIR can be used in multi-user system. The E-classroom using MCJITIR is of great meaning to enhance initiative learning.
In future work, the design need be improved based on evaluation of performance. In the method, the media sense deserves research. Sense of text and media may be integrated. To application fields, digital library and digital museum can use MCJITIR. The other mental contents in MCJITIR need be studied and explored further.
引文
[1]Bradley James Rhodes.Using Physical Context for Just-In-Time Information Retrieval[J].IEEE TRANSACTIONS ON COMPUTERS,2003.8,VOL.52.NO.8:1011--1014.
[2]Bradley James Rhodes.Just-In-Time Information Retrieval[DOCTORAL DISSERTATION][D],MIT,2000.
[3]杨炳儒,李晋宏,宋威等.面向复杂系统的知识发现过程模型KD(D&K)及其应用[J].自动化学报,2007.2,Vol.37.No.2:151-155.
[4]YANG Bingru,ZHANG Taohong.Coordinators based on cognitive psychology features and the corresponding KDD process model(in English)[J].中国科学技术大学学报,2007.2,Vol.37.No.2:212-216.
[5]Tao Li,Chang-shing Perng.KDD2006 Workshop Report Theory and Practice of Temporal Data Mining[J].ACM SIGKDD Explorations.2006,Vol.8.No.2:96-97.
[6]LanH.Witten.数据挖掘:实用机器学习技术(英文影印版·第2版)[M].北京:机械工业出版社,2005:253-267
[7]杨炳儒 著.基于内在基理的知识发现理论及其应用[M].北京:电子工业出版社,2004:78-91,
[8]Michel A.Rbeiz.Semantic Representation of Digital Ink in the Classroom Learning PaTter[D].MIT,2006
[9]王汝传,徐小龙,黄海平编著.智能Agent及其在信息网络中的应用[M].北京:北京邮电大学出版社,2006:43-69
[10]UMBC Agent Web.http://agents,umbc.edu/[OL].[2007-09-10].
[11]Mani Subramanian.NETWORK MANAGEMENT:Principles and Practice[M].Higher Education Press&Pearson Education,2001:105-109
[12]丁永生 编著.计算智能[M].北京:科学出版社,2004:325-359
[13]沈固朝 主编.网络信息检索:工具·方法·实践[M].北京:高等教育出版社,2004:100-152
[14]蔡自兴,姚莉著.人工智能[M].长沙:国防科技大学出版社,2006:99-107
[15]毛国君,段立娟,王实等编著.数据挖掘原理与算法[M].北京:清华大学出版社,2005:211-249
[16]朱三元,钱乐秋,宿为民编著.软件工程技术概论[M].北京:科学出版社,2002:49-94
[17]孙建军,成颖 编著.信息检索技术[M].北京:科学出版社,2004:427-458
[18]康师友 编著.人工智能技术导论[M].西安:西安电子科技大学出版社,2002:263-267
[19]柯平 主编.信息素养与信息检索概论[M].天津:南开大学出版社,2005:40-44
[20]毛新军 编著.面向主体的软件开发[M].北京:清华大学出版社,2005:106-117
[21]Andrew.COMPUTER NETWORK(FOURTH EDITION)(影印版)[M].清华大学出版社,2004:819-826
[22]Hua Xiang.A Computational.Model and Psychological Experiment Analysis on Affective Information Processing[DOCTORAL DISSERTATION][D],The University of Tokushima,2007
[23]陈世平,丁山山,胡成梅等.个性化Web信息代理的研究与开发[J].上海理工大学学报,2004.6,Vol.26.No.6:575-579
[24]胡军.面向自治计算的基于政策的多agent协同体系研究[博士学位论文][D].杭州:浙江大学计算机科学与技术学院,2006
[25]范明,范宏建等译.数据挖掘导论[M].北京:人民邮电出版社,2006:201-214
[26]章兢,张小刚等编著.数据挖掘算法及其工程应用[M].北京:机械工业出版社,2006:1-10
[27]David Konopnicki,Oded Shmueli.Database-Inspired Search[C].Trondheim,Norway:Proceedings of the 31st VLDB Conference,2005:1-12
[28]Agents.http://www.aaai.org/AITopics/pmwiki/pmwiki,php/AITopics /Agents[OL][2008-03-05]
[29]王珑,章毅,吴跃.MAS环境下实现Agent交互协作的关键技术[J].电子科技大学学报,2003.2,Vol.32.No.2:158-163
[30]刘海庆,邹云松,刘孟仁.基于Agent的“信息路由”通信模型[J].海军工程大学学报,2004.5,Vol.16.No.5:85-90
[31]罗建,王成良.基于KQML的Agent通信体系结构设计[J].重庆大学学报,2004.3.Vol.27.No.3:56-60
[32]S.Brin,L Page.The anatomy of a large-scale hypertexitual Web search engine[C],the 7th Int.Conf.WWW,Brisbane.Australia,April,1998
[33]王汝传,赵新宁,王绍棣.基于网络的移动代理系统安全模型的研究和分析[J].计算机学报,2003.4,26(4):477-483
[34]庄超.一种新型Internet内容版权保护的计算机制[J].计算机学报,2000.10,vol.23.No.10:1088-1091
[35]中国计算机学会学术工作委员会.中国计算机科学技术发展报告(2005)[M].北京:清华大学出版社,2006:91-109
[36]史忠植 著.知识发现[M].北京:清华大学出版社,2002:13-20
[37]卢亮,张博文编著.搜索引擎——原理、实践与应用[M].北京:电子工业出版社,2007:10-111
[38]李晓明,闫宏飞,王继民编著.搜索引擎——原理、技术与系统[M].北京:电子工业出版社,2007:187-207
[39]Grady Booch,James Rumbaugh,Ivar Jacobson.The unified modeling language user guide[M].China Machine Press,2006:103-114
[40]冷彪,覃征.基于先验知识的三维模型特征选择算法[J].清华大学学报(自然科学版),2008.4,vol.48.No.4:586-588
[41]MITOPENCOURSEWARE.http://ocw.mit.edu/OcwWeb/web/home/home/index.htm:[OL].[2008-01-30]
[42]大学课程在线.http://realcourse.grids.cn/[OL].[2008-02-27]
[43]郎皓,王斌,李锦涛等.文本检索的查询性能评测[J].软件学报,2008.vol.19No.1:48-61
[44]http://porta1.acm.org/citation.cfm?id=1011437[OL].[2007-11-18]
[45]Zhe Wang,Wei Dong,William Josephson,et al..Sizing Sketches:A Rank-Based Analysis for Similarity Search[C].SIGMETRICS'07,2007,SanDiego,California,USA.
[46]The Content-Aware Similarity Search.http://www.cs.princeton.edu/cass [OL].[2008-01-20]
[47]陈红刚,庄超.基于数据和知识协同的即时信息检索[J].计算机与数字工程,2008.4,vol.36.No.4:151-153
[48]陈红刚,庄超.基于多协同的即时信息检索框架[J].现代图书情报技术,2008.2,vol.19.No.2:48-52
[2]Bradley James Rhodes.Just-In-Time Information Retrieval[DOCTORAL DISSERTATION][D],MIT,2000.
[3]杨炳儒,李晋宏,宋威等.面向复杂系统的知识发现过程模型KD(D&K)及其应用[J].自动化学报,2007.2,Vol.37.No.2:151-155.
[4]YANG Bingru,ZHANG Taohong.Coordinators based on cognitive psychology features and the corresponding KDD process model(in English)[J].中国科学技术大学学报,2007.2,Vol.37.No.2:212-216.
[5]Tao Li,Chang-shing Perng.KDD2006 Workshop Report Theory and Practice of Temporal Data Mining[J].ACM SIGKDD Explorations.2006,Vol.8.No.2:96-97.
[6]LanH.Witten.数据挖掘:实用机器学习技术(英文影印版·第2版)[M].北京:机械工业出版社,2005:253-267
[7]杨炳儒 著.基于内在基理的知识发现理论及其应用[M].北京:电子工业出版社,2004:78-91,
[8]Michel A.Rbeiz.Semantic Representation of Digital Ink in the Classroom Learning PaTter[D].MIT,2006
[9]王汝传,徐小龙,黄海平编著.智能Agent及其在信息网络中的应用[M].北京:北京邮电大学出版社,2006:43-69
[10]UMBC Agent Web.http://agents,umbc.edu/[OL].[2007-09-10].
[11]Mani Subramanian.NETWORK MANAGEMENT:Principles and Practice[M].Higher Education Press&Pearson Education,2001:105-109
[12]丁永生 编著.计算智能[M].北京:科学出版社,2004:325-359
[13]沈固朝 主编.网络信息检索:工具·方法·实践[M].北京:高等教育出版社,2004:100-152
[14]蔡自兴,姚莉著.人工智能[M].长沙:国防科技大学出版社,2006:99-107
[15]毛国君,段立娟,王实等编著.数据挖掘原理与算法[M].北京:清华大学出版社,2005:211-249
[16]朱三元,钱乐秋,宿为民编著.软件工程技术概论[M].北京:科学出版社,2002:49-94
[17]孙建军,成颖 编著.信息检索技术[M].北京:科学出版社,2004:427-458
[18]康师友 编著.人工智能技术导论[M].西安:西安电子科技大学出版社,2002:263-267
[19]柯平 主编.信息素养与信息检索概论[M].天津:南开大学出版社,2005:40-44
[20]毛新军 编著.面向主体的软件开发[M].北京:清华大学出版社,2005:106-117
[21]Andrew.COMPUTER NETWORK(FOURTH EDITION)(影印版)[M].清华大学出版社,2004:819-826
[22]Hua Xiang.A Computational.Model and Psychological Experiment Analysis on Affective Information Processing[DOCTORAL DISSERTATION][D],The University of Tokushima,2007
[23]陈世平,丁山山,胡成梅等.个性化Web信息代理的研究与开发[J].上海理工大学学报,2004.6,Vol.26.No.6:575-579
[24]胡军.面向自治计算的基于政策的多agent协同体系研究[博士学位论文][D].杭州:浙江大学计算机科学与技术学院,2006
[25]范明,范宏建等译.数据挖掘导论[M].北京:人民邮电出版社,2006:201-214
[26]章兢,张小刚等编著.数据挖掘算法及其工程应用[M].北京:机械工业出版社,2006:1-10
[27]David Konopnicki,Oded Shmueli.Database-Inspired Search[C].Trondheim,Norway:Proceedings of the 31st VLDB Conference,2005:1-12
[28]Agents.http://www.aaai.org/AITopics/pmwiki/pmwiki,php/AITopics /Agents[OL][2008-03-05]
[29]王珑,章毅,吴跃.MAS环境下实现Agent交互协作的关键技术[J].电子科技大学学报,2003.2,Vol.32.No.2:158-163
[30]刘海庆,邹云松,刘孟仁.基于Agent的“信息路由”通信模型[J].海军工程大学学报,2004.5,Vol.16.No.5:85-90
[31]罗建,王成良.基于KQML的Agent通信体系结构设计[J].重庆大学学报,2004.3.Vol.27.No.3:56-60
[32]S.Brin,L Page.The anatomy of a large-scale hypertexitual Web search engine[C],the 7th Int.Conf.WWW,Brisbane.Australia,April,1998
[33]王汝传,赵新宁,王绍棣.基于网络的移动代理系统安全模型的研究和分析[J].计算机学报,2003.4,26(4):477-483
[34]庄超.一种新型Internet内容版权保护的计算机制[J].计算机学报,2000.10,vol.23.No.10:1088-1091
[35]中国计算机学会学术工作委员会.中国计算机科学技术发展报告(2005)[M].北京:清华大学出版社,2006:91-109
[36]史忠植 著.知识发现[M].北京:清华大学出版社,2002:13-20
[37]卢亮,张博文编著.搜索引擎——原理、实践与应用[M].北京:电子工业出版社,2007:10-111
[38]李晓明,闫宏飞,王继民编著.搜索引擎——原理、技术与系统[M].北京:电子工业出版社,2007:187-207
[39]Grady Booch,James Rumbaugh,Ivar Jacobson.The unified modeling language user guide[M].China Machine Press,2006:103-114
[40]冷彪,覃征.基于先验知识的三维模型特征选择算法[J].清华大学学报(自然科学版),2008.4,vol.48.No.4:586-588
[41]MITOPENCOURSEWARE.http://ocw.mit.edu/OcwWeb/web/home/home/index.htm:[OL].[2008-01-30]
[42]大学课程在线.http://realcourse.grids.cn/[OL].[2008-02-27]
[43]郎皓,王斌,李锦涛等.文本检索的查询性能评测[J].软件学报,2008.vol.19No.1:48-61
[44]http://porta1.acm.org/citation.cfm?id=1011437[OL].[2007-11-18]
[45]Zhe Wang,Wei Dong,William Josephson,et al..Sizing Sketches:A Rank-Based Analysis for Similarity Search[C].SIGMETRICS'07,2007,SanDiego,California,USA.
[46]The Content-Aware Similarity Search.http://www.cs.princeton.edu/cass [OL].[2008-01-20]
[47]陈红刚,庄超.基于数据和知识协同的即时信息检索[J].计算机与数字工程,2008.4,vol.36.No.4:151-153
[48]陈红刚,庄超.基于多协同的即时信息检索框架[J].现代图书情报技术,2008.2,vol.19.No.2:48-52