基于多Agent的智能网络教学系统模型研究
摘要
随着Internet的发展,网络教学得到了越来越广泛的应用,人们可以不受时间和空间的限制自主地参加网上学习。这种学习方式能够激发学习者的学习兴趣,从而达到让学习者主动构建知识的目的,实现自己获取知识、甚至创新知识的理想目标。因此,网络教学系统也成为了国内外教育技术领域的研究热点。但是目前的网络教学系统还存在着缺乏先进的学习理论的指导、系统设计缺乏智能性等缺点。
论文针对现有网络教学系统存在的不足,引用Agent技术,设计了一个基于多Agent的网络教学系统模型。该模型分三个模块,各个模块由多个Agent组成。在论文中分别对这些模块的结构和功能进行了分析。模型的设计思想即发挥了教师的主导作用,又充分体现了学生的主体作用。
系统的智能化设计方面,在知识表示方法上,为了对学科内容提供全局的认识,采用树结构来组织课程内容知识点,通过对知识点属性赋值建立知识点之间的语义联系,为有效控制教学序列提供依据。提出了基于错误诊断的强化学习策略,通过分析测试答案找出学生的知识缺陷,结合知识表示模型,推导出学生需要加强的合理学习路径,达到能对学生进行有针对性的强化教学的目的。提出了能够根据学生的认知水平自适应的进行组卷的智能组卷算法。针对现有学生模型的不足,构造了一个基于认知理论的学生模型,提出了评估认知水平、记录学习进度、估算学习偏好的方法。
通过对多Agent和J2EE技术的深入学习和研究,提出了集成JADE和J2EE来实现智能网络教学模型的思想。该系统模型在一定程度上实现了智能化、个性化教学,可以为学习者提供更方便和更高效的网络学习环境。最后以《Java程序设计》课程为例给出了该系统的实验结果,验证了该系统的有效性和实用性。
With the development of Internet, network teaching has been getting more and more extensive application, students can participate in online learning free and not restricting by time and space. This learning means can inspire learners to learn, so that learners can construct knowledge depending on their own initiatively and get, even innovate knowledge by themselves. As a result, the network teaching system has become the the research heat of educational technology field at home and abroad. However, there are still many shortcomings in the network teaching system nowadays, such as lack guidance of the advanced learning theory and intelligence.
Aimed at defects of the present network teaching system, I design an intelligent network teaching system, which is based on Agent technology. The model is divided into three modules, each module consists of a number of Agent. In this thesis, I analyze the structure and function of each module. The designing concept of the model not only bring a leading role of teacher into play but also fully reflects the main body of student.
In the aspects of designing intelligent system, in aspect of knowledge expression, in order to offer the overall view of discipline content, the tree structure is used to organize knowledge points, by setting some knowledge node attribute the semantic relation between knowledge nodes is established to offer the basis to control teaching sequence effectively. A strengthened study strategy based on fault diagnosis is proposed. Test results are analyzed to find students' knowledge flaws, and according to the knowledge expression model a proper study path needed to be strengthened is deduced to achieve the goal to provide pertinent strengthened teaching for students. According to the students’level of learning, an intelligent algorithm for test paper is proposed. Aiming at the shortcoming of the existing student models, a student model that can reflect cognitive ability is constructed. The method of evaluating cognitive ability, recording study progress and computing mood in study is proposed.
With the research on Multi-Agent and J2EE technology, the thesis realized the system based on JADE and J2EE. To some extent this network teaching system realizes the intelligent teaching and provides more convenient and highly effective network-learning environment for learners. Finally, "Java programming" courses as an example of the system provides the results of system experimentation and verifies the effectivity and practicality.
论文针对现有网络教学系统存在的不足,引用Agent技术,设计了一个基于多Agent的网络教学系统模型。该模型分三个模块,各个模块由多个Agent组成。在论文中分别对这些模块的结构和功能进行了分析。模型的设计思想即发挥了教师的主导作用,又充分体现了学生的主体作用。
系统的智能化设计方面,在知识表示方法上,为了对学科内容提供全局的认识,采用树结构来组织课程内容知识点,通过对知识点属性赋值建立知识点之间的语义联系,为有效控制教学序列提供依据。提出了基于错误诊断的强化学习策略,通过分析测试答案找出学生的知识缺陷,结合知识表示模型,推导出学生需要加强的合理学习路径,达到能对学生进行有针对性的强化教学的目的。提出了能够根据学生的认知水平自适应的进行组卷的智能组卷算法。针对现有学生模型的不足,构造了一个基于认知理论的学生模型,提出了评估认知水平、记录学习进度、估算学习偏好的方法。
通过对多Agent和J2EE技术的深入学习和研究,提出了集成JADE和J2EE来实现智能网络教学模型的思想。该系统模型在一定程度上实现了智能化、个性化教学,可以为学习者提供更方便和更高效的网络学习环境。最后以《Java程序设计》课程为例给出了该系统的实验结果,验证了该系统的有效性和实用性。
With the development of Internet, network teaching has been getting more and more extensive application, students can participate in online learning free and not restricting by time and space. This learning means can inspire learners to learn, so that learners can construct knowledge depending on their own initiatively and get, even innovate knowledge by themselves. As a result, the network teaching system has become the the research heat of educational technology field at home and abroad. However, there are still many shortcomings in the network teaching system nowadays, such as lack guidance of the advanced learning theory and intelligence.
Aimed at defects of the present network teaching system, I design an intelligent network teaching system, which is based on Agent technology. The model is divided into three modules, each module consists of a number of Agent. In this thesis, I analyze the structure and function of each module. The designing concept of the model not only bring a leading role of teacher into play but also fully reflects the main body of student.
In the aspects of designing intelligent system, in aspect of knowledge expression, in order to offer the overall view of discipline content, the tree structure is used to organize knowledge points, by setting some knowledge node attribute the semantic relation between knowledge nodes is established to offer the basis to control teaching sequence effectively. A strengthened study strategy based on fault diagnosis is proposed. Test results are analyzed to find students' knowledge flaws, and according to the knowledge expression model a proper study path needed to be strengthened is deduced to achieve the goal to provide pertinent strengthened teaching for students. According to the students’level of learning, an intelligent algorithm for test paper is proposed. Aiming at the shortcoming of the existing student models, a student model that can reflect cognitive ability is constructed. The method of evaluating cognitive ability, recording study progress and computing mood in study is proposed.
With the research on Multi-Agent and J2EE technology, the thesis realized the system based on JADE and J2EE. To some extent this network teaching system realizes the intelligent teaching and provides more convenient and highly effective network-learning environment for learners. Finally, "Java programming" courses as an example of the system provides the results of system experimentation and verifies the effectivity and practicality.
引文
[1]倪宇斌.基于Agent的网络教学系统模型的研究与设计[D]: [硕士研究生论文].广州:华南师范大学计算机学院, 2007.
[2]李海燕,费洪晓.基于多Agent的智能网络教学系统研究[J].南华大学学报(自然科学版), 2004, Vol.18(4): 71-74.
[3]常守金.计算机辅助教育简明教程[M].天津:天津科学技术出版社, 1994: 161-167.
[4]王洪亚,宁国勤,陆三兰.基于Agent的网上协作学习环境的研究[J].鄂州大学报, 2001, 8(4): 29-31.
[5]冯永杰,孟宾,翟玉庆. Agent在智能信息检索中的应用研究[J].计算机应用研究, 2002, 19(2): 33-35.
[6]万江平.用Java语言开发专家系统[J].计算机应用研究, 2000, 17(5): 62-64.
[7]史忠植,王文杰,太启家.智能主体研究现状和发展趋势[J].计算机世界(周报),技术专题版, 1998(4).
[8] W. Lewis Johnson and Erin Shaw. Using Agents to Overcome Deficiencies in Web-Based Courseware[EB/OL]. http://www.isi.edu/isd/johnson.html, 2003-03.
[9] Claude Fransson, Thierry Mengelle, Esma Aimeur. Using Pedagogical Agents In a Multi-Strategic Intelligent Tutoring System[C]. Proceedings of AI-ED, 1997.
[10] Evandro de Barros Costa. A Multi-Agent Interactive Learning Environment Model[C]. Proceedings of ICCE97.
[11] Jean-Francois Morin, Ruddy Lelouche. Agent-oriented tutoring knowledge modeling in a problem-solving ITS [EB/OL]. http://www.ift.ulaval.ca/~lelouche/Publications/Morin_ Lelouche_ 199 8InteractionAgents.pdf, 1997.
[12]杨锦潭,伍石能.仿真教学代理人最新发展与挑战[EB/OL]. http://ia.nknu.edu.tw/10.pdf, 2004.
[13]刘树勋,李卫华,区益善. Internet网上智能检索Agent的信息检索方法[J].广东工业大学学报,1999, 15(4): 39-42.
[14]李海燕.基于多Agent的智能网络教学系统研究[D]: [硕士研究生论文].长沙:中南大学, 2004.
[15]李广,姜英杰.个性化学习的理论建构与特征分析[J].东北师大学报, 2005(3): 152-156
[16] David Griffiths. Environmental Challenges:Making a Difference in the Classroom[M]. Proceedings of CAL 97, 1997, 95-99.
[17]何克杭.建构主义学习理论与建构主义学习环境[J].教育传播与技术, 1996, 12 (3): 15-20.
[18]王超.建构主义学习理论对教学的启示[J].黑龙江高教研究, 2006, 7: 111-112.
[19]李艳.基于多Agent系统的远程协同教研平台的研究与实现[D]: [硕士研究生论文].福州:福州大学数学与计算机科学学院, 2005.
[20] Nwana H.S. Software agents: an overview[J]. The Knowledge Engineering Review, 1997, 58(2): 205-244.
[21] Y. Shoham. Agent-Oriented Programming[J]. Artificial Intelligence, 1993, 50: 51-92.
[22] Wooldridge M, Jennings N. Intelligent Agents: theory and practice[J]. The knowledge Engineering Review, 1995, 10(2): 115~152.
[23] Wooldfidge M. An introduction to multiagent systems[M].北京:电子工业出版社, 2003.
[24]刘勇.多Agent系统理论和应用研究[D]: [博士学位论文].重庆:重庆大学计算机科学与工程学院, 2003.
[25]史忠植.智能主体及应用[M].科学出版社, 2000.
[26]彭敏佳.基于Agent的网络教学系统的研究与应用[D]: [硕士研究生论文].重庆:重庆大学计算机学院, 2004.
[27]李海刚,吴启迪.多Agent系统研究综述[J].同济大学学报, 2003, 31(6): 728-731.
[28]倪敏.基于Agent的网络教学系统设计与实现[D]: [硕士研究生论文].长沙:中南大学, 2007.
[29]王汝传,徐小龙,黄海平.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, 286-305.
[30]陈桂香,周兴斌.一种面向Agent的智能大系统[J].南昌大学学报(理科版), 2000, 12.
[31]孙志勇.多Agent系统体系结构及建模方法研究[D]: [博士学位论文].合肥:合肥工业大学,2004, 6.
[32] Momica Pawlan. Writing Enterprise Applications with Java 2 Enterprise Edition[M]. 2000, 6.
[33]邢国庆等译. J2EE应用[M].北京:电子工业出版社, 2002.
[34]石鸥燕.基于J2EE的多Agent网络教学系统的研究与实现[D]: [硕士研究生论文].天津:天津师范大学教育技术学, 2005.
[35]冯燕奎等. JSP实用案例教程[M].北京:清华大学出版社, 2004.
[36]杨学瑜,王志军,刘同利. JSP入门与提高[M].北京:清华大学出版社,2002,3(1), 5-7.
[37]谢祥志.基于Web和Agent的智能实验教学系统的研究与实现[D]: [硕士研究生论文].西安:西安科技大学计算机应用技术, 2006.
[38]李繁.基于B/S模式的网上教学系统的设计与实现[D]: [硕士研究生论文].大连:大连理工大学, 2005.
[39]段培俊,周东岱,程晓春.基于Multi-Agent技术的教学资源个性化服务模型研究[J].东北师大学报, 2006, 38 (2) : 31-35.
[40]刘莹.基于多Agent的网络教学系统的研究[D]: [硕士研究生论文].南京:河海大学信息工程学院, 2006.
[41]童红霞,谢深泉. ICAI中知识点关系的研究[J].计算机工程与应用, 2004, 1: 77-78.
[42]程蕾.基于Agent技术的智能题库系统的研究与设计[D]: [硕士研究生论文].杭州:浙江大学, 2003, 36-47.
[43]岳永恒.基于Agent教学系统中关键问题的研究[D]: [硕士研究生论文].长沙:长沙理工大学计算机与通信工程学院, 2005.
[44]王立新.自适应模糊系统与控制[M].北京:国防工业出版社, 1995: 241-245.
[45]王士同.神经模糊系统及其应用[M].北京:北京航空航天大学出版社, 1998, 6.
[46] H-J, Zimmermann.模糊集决策和专家系统[M].石家庄:河北教育出版社, 1992.
[47] Katz. S., Lesgold.A.Eggan. Gordin, M.. Modeling the student in Sherlock II[J]. Journal of Artificial Intelligence in Education 3(4): 495-518.
[48] Gurer.D.desJardins, M.Schlager.M. Representing a student's learning states and transitions[J]. American Association of Artificial Intelligence Spring Symposium on Representing Mental States and Mechanisms, 1995.
[49]何炎祥,陈莘明. Agent和Agent系统的设计与应用[M].武汉:武汉大学出版社,2001.
[50]彭庆波.基于多Agent的智能网络教学系统研究[D]: [硕士研究生论文].长沙:国防科学技术大学管理科学与工程, 2005.
[51] Fabio Bellifemine, Giovanni Caire, Tiziana Trucco. JADE PROGRAMMER’S GUIDE[EB/OL]. http:// jade.cselt.it, 2004-07.
[52] Fabio Bellifemine, Giovanni Caire, Dominic Greenwood. Developing Multi-Agent Systems with JADE[M]. 2006, 15-129.
[53] Giovanni Caire. JADE PROGRAMMING FOR BEGINNERS[EB/OL]. http://jade.cselt.it, 2003-12.
[54]金明宇,翟玉庆.基于Web的Agent通信框架[J].计算机工程, 2003, 29 (3): 195-197.
[55] Jean Vaucher, Ambroise Ncho. JADE Tutorial and Primer[M]. 2003-09.
[56] Fabio Bellifemine, Agostino Poggi, GiovanniRimassa. JADE-A FIPA-compliant Framework [EB/OL]. http://jade.cselt.it, 2000.
[57] Jean Vaucher, Ambroise Ncho. JADE Tutorial and Primer[M]. 2003-09.
[58] Dominic G. FIPA-The Foundation for Intelligent Physical Agents[EB/OL]. http://jade.cselt.it, 2004, 07.
[59] Giovanni Caire. JADE: The new kernel and last developments[EB/OL].http://JADE.blab.com/papers /JADE-the-services-architecture.pdf, 2004-09.
[60] C.J. Butz, S. Hua, R.B. Maguire.A web-based bayesian intelligent tutoring system for computer programming. Web Intelligence and Agent Systems[J]. 2006, 4 (1)77-97.
[2]李海燕,费洪晓.基于多Agent的智能网络教学系统研究[J].南华大学学报(自然科学版), 2004, Vol.18(4): 71-74.
[3]常守金.计算机辅助教育简明教程[M].天津:天津科学技术出版社, 1994: 161-167.
[4]王洪亚,宁国勤,陆三兰.基于Agent的网上协作学习环境的研究[J].鄂州大学报, 2001, 8(4): 29-31.
[5]冯永杰,孟宾,翟玉庆. Agent在智能信息检索中的应用研究[J].计算机应用研究, 2002, 19(2): 33-35.
[6]万江平.用Java语言开发专家系统[J].计算机应用研究, 2000, 17(5): 62-64.
[7]史忠植,王文杰,太启家.智能主体研究现状和发展趋势[J].计算机世界(周报),技术专题版, 1998(4).
[8] W. Lewis Johnson and Erin Shaw. Using Agents to Overcome Deficiencies in Web-Based Courseware[EB/OL]. http://www.isi.edu/isd/johnson.html, 2003-03.
[9] Claude Fransson, Thierry Mengelle, Esma Aimeur. Using Pedagogical Agents In a Multi-Strategic Intelligent Tutoring System[C]. Proceedings of AI-ED, 1997.
[10] Evandro de Barros Costa. A Multi-Agent Interactive Learning Environment Model[C]. Proceedings of ICCE97.
[11] Jean-Francois Morin, Ruddy Lelouche. Agent-oriented tutoring knowledge modeling in a problem-solving ITS [EB/OL]. http://www.ift.ulaval.ca/~lelouche/Publications/Morin_ Lelouche_ 199 8InteractionAgents.pdf, 1997.
[12]杨锦潭,伍石能.仿真教学代理人最新发展与挑战[EB/OL]. http://ia.nknu.edu.tw/10.pdf, 2004.
[13]刘树勋,李卫华,区益善. Internet网上智能检索Agent的信息检索方法[J].广东工业大学学报,1999, 15(4): 39-42.
[14]李海燕.基于多Agent的智能网络教学系统研究[D]: [硕士研究生论文].长沙:中南大学, 2004.
[15]李广,姜英杰.个性化学习的理论建构与特征分析[J].东北师大学报, 2005(3): 152-156
[16] David Griffiths. Environmental Challenges:Making a Difference in the Classroom[M]. Proceedings of CAL 97, 1997, 95-99.
[17]何克杭.建构主义学习理论与建构主义学习环境[J].教育传播与技术, 1996, 12 (3): 15-20.
[18]王超.建构主义学习理论对教学的启示[J].黑龙江高教研究, 2006, 7: 111-112.
[19]李艳.基于多Agent系统的远程协同教研平台的研究与实现[D]: [硕士研究生论文].福州:福州大学数学与计算机科学学院, 2005.
[20] Nwana H.S. Software agents: an overview[J]. The Knowledge Engineering Review, 1997, 58(2): 205-244.
[21] Y. Shoham. Agent-Oriented Programming[J]. Artificial Intelligence, 1993, 50: 51-92.
[22] Wooldridge M, Jennings N. Intelligent Agents: theory and practice[J]. The knowledge Engineering Review, 1995, 10(2): 115~152.
[23] Wooldfidge M. An introduction to multiagent systems[M].北京:电子工业出版社, 2003.
[24]刘勇.多Agent系统理论和应用研究[D]: [博士学位论文].重庆:重庆大学计算机科学与工程学院, 2003.
[25]史忠植.智能主体及应用[M].科学出版社, 2000.
[26]彭敏佳.基于Agent的网络教学系统的研究与应用[D]: [硕士研究生论文].重庆:重庆大学计算机学院, 2004.
[27]李海刚,吴启迪.多Agent系统研究综述[J].同济大学学报, 2003, 31(6): 728-731.
[28]倪敏.基于Agent的网络教学系统设计与实现[D]: [硕士研究生论文].长沙:中南大学, 2007.
[29]王汝传,徐小龙,黄海平.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, 286-305.
[30]陈桂香,周兴斌.一种面向Agent的智能大系统[J].南昌大学学报(理科版), 2000, 12.
[31]孙志勇.多Agent系统体系结构及建模方法研究[D]: [博士学位论文].合肥:合肥工业大学,2004, 6.
[32] Momica Pawlan. Writing Enterprise Applications with Java 2 Enterprise Edition[M]. 2000, 6.
[33]邢国庆等译. J2EE应用[M].北京:电子工业出版社, 2002.
[34]石鸥燕.基于J2EE的多Agent网络教学系统的研究与实现[D]: [硕士研究生论文].天津:天津师范大学教育技术学, 2005.
[35]冯燕奎等. JSP实用案例教程[M].北京:清华大学出版社, 2004.
[36]杨学瑜,王志军,刘同利. JSP入门与提高[M].北京:清华大学出版社,2002,3(1), 5-7.
[37]谢祥志.基于Web和Agent的智能实验教学系统的研究与实现[D]: [硕士研究生论文].西安:西安科技大学计算机应用技术, 2006.
[38]李繁.基于B/S模式的网上教学系统的设计与实现[D]: [硕士研究生论文].大连:大连理工大学, 2005.
[39]段培俊,周东岱,程晓春.基于Multi-Agent技术的教学资源个性化服务模型研究[J].东北师大学报, 2006, 38 (2) : 31-35.
[40]刘莹.基于多Agent的网络教学系统的研究[D]: [硕士研究生论文].南京:河海大学信息工程学院, 2006.
[41]童红霞,谢深泉. ICAI中知识点关系的研究[J].计算机工程与应用, 2004, 1: 77-78.
[42]程蕾.基于Agent技术的智能题库系统的研究与设计[D]: [硕士研究生论文].杭州:浙江大学, 2003, 36-47.
[43]岳永恒.基于Agent教学系统中关键问题的研究[D]: [硕士研究生论文].长沙:长沙理工大学计算机与通信工程学院, 2005.
[44]王立新.自适应模糊系统与控制[M].北京:国防工业出版社, 1995: 241-245.
[45]王士同.神经模糊系统及其应用[M].北京:北京航空航天大学出版社, 1998, 6.
[46] H-J, Zimmermann.模糊集决策和专家系统[M].石家庄:河北教育出版社, 1992.
[47] Katz. S., Lesgold.A.Eggan. Gordin, M.. Modeling the student in Sherlock II[J]. Journal of Artificial Intelligence in Education 3(4): 495-518.
[48] Gurer.D.desJardins, M.Schlager.M. Representing a student's learning states and transitions[J]. American Association of Artificial Intelligence Spring Symposium on Representing Mental States and Mechanisms, 1995.
[49]何炎祥,陈莘明. Agent和Agent系统的设计与应用[M].武汉:武汉大学出版社,2001.
[50]彭庆波.基于多Agent的智能网络教学系统研究[D]: [硕士研究生论文].长沙:国防科学技术大学管理科学与工程, 2005.
[51] Fabio Bellifemine, Giovanni Caire, Tiziana Trucco. JADE PROGRAMMER’S GUIDE[EB/OL]. http:// jade.cselt.it, 2004-07.
[52] Fabio Bellifemine, Giovanni Caire, Dominic Greenwood. Developing Multi-Agent Systems with JADE[M]. 2006, 15-129.
[53] Giovanni Caire. JADE PROGRAMMING FOR BEGINNERS[EB/OL]. http://jade.cselt.it, 2003-12.
[54]金明宇,翟玉庆.基于Web的Agent通信框架[J].计算机工程, 2003, 29 (3): 195-197.
[55] Jean Vaucher, Ambroise Ncho. JADE Tutorial and Primer[M]. 2003-09.
[56] Fabio Bellifemine, Agostino Poggi, GiovanniRimassa. JADE-A FIPA-compliant Framework [EB/OL]. http://jade.cselt.it, 2000.
[57] Jean Vaucher, Ambroise Ncho. JADE Tutorial and Primer[M]. 2003-09.
[58] Dominic G. FIPA-The Foundation for Intelligent Physical Agents[EB/OL]. http://jade.cselt.it, 2004, 07.
[59] Giovanni Caire. JADE: The new kernel and last developments[EB/OL].http://JADE.blab.com/papers /JADE-the-services-architecture.pdf, 2004-09.
[60] C.J. Butz, S. Hua, R.B. Maguire.A web-based bayesian intelligent tutoring system for computer programming. Web Intelligence and Agent Systems[J]. 2006, 4 (1)77-97.