基于本体论的玉米病虫害诊断系统的设计与实现
摘要
目前的知识系统大多是按照用户的不同需求,采用不同的知识表示方法、不同的建模方法、不同的开发工具,在不同的平台上进行开发或者二次开发。在知识系统之间实现互操作十分困难,更不用说协同各系统的功能去解决问题了。语义网和本体论的相关研究和应用,为解决分布的、异构的知识系统在知识共享、复用和互操作等方面的困难,提供了相应的技术基础和可行的解决方案。本体为给定的关注域给出了一个共同理解的明确说明来促进各种主体之间的交流,通过这种交流来实现给定域中知识在不同主体之间的复用和共享。通过建立领域本体来表达领域知识,为专家系统建立可共享、可重用的知识库提供参考,进而促进系统间的互操作并降低系统开发成本。因此,构建基于本体的专家系统成为了新的研究方向。
本文对基于本体论的专家系统的构建进行了研究,具体工作包括构建领域本体和设计基于本体的专家系统两部分。本文的工作属于国家863项目“数字农业知识网格技术研究及应用”中“基于Web的分布式农业生产知识咨询平台”的一部分。
本文具体内容：
1.提出了一个领域本体的构建方法——“九步法”。
对本体的概念、主要的本体描述语言、构建方法及开发工具等理论知识进行了研究,提出了一个领域本体的构建方法——“九步法”。“九步法”包括以下九步：确定本体的领域和范围；考虑复用已有本体；获取知识；列举术语；定义类和层次；定义属性；定义侧面；定义实例；检查异常。
2.依照“九步法”实施本体工程,构建了玉米病害诊断本体和玉米虫害诊断本体。
在构建本体时参考了相关农学资料,并接受了农学专家的指导。通过建立概念、概念间的关系、定义实例、定义属性的约束等方式表达了玉米病虫害诊断的领域知识。该本体采用了W3C推荐的本体描述语言OWL进行描述,利用斯坦福大学开发的本体构建工具Protege3.4.4进行开发。
3.利用构建的玉米病害诊断本体和玉米虫害诊断本体作为玉米病虫害诊断系统的知识库,采用Visual C# 2008和Microsoft SQL Server 2005开发了基于“图像规则”的玉米病虫害诊断系统。
从玉米病虫害防治工作的实际情况出发,进行了系统的需求分析和设计,明确了系统的用户、功能和框架。系统采用B/S结构,按照功能划分为知识层、数据层、逻辑控制层和人机交互层。系统的功能包括对玉米常见病虫害的诊断,病虫害详细信息及对应防治方案的查询。系统不仅能对当前已经发生的病虫害进行诊断,也能将潜在的虫害威胁反馈给用户；系统还提供了多种诊断模式,支持用户根据生产中的实际情况选择合适的模式进行诊断。
本文工作的意义：
本文提出的“九步法”适用于由知识工程师构建领域本体的场景,可以较好的支持本体的维护和进化工作；通过构建玉米病害诊断本体以及玉米虫害诊断本体,为玉米病虫害诊断专家系统建立可共享、可重用的知识库提供了参考；基于本体的玉米病虫害诊断系统不但美观易用,而且满足了用户在农业生产中对信息服务的实际需求,具有很强的实用性。
Most of the current knowledge systems are developed or redeveloped base on the user's different needs, using different knowledge representation methods, different modeling methods and different development tools. Therefore, the interoperability between knowledge systems becomes a problem. not to mention coordinate these systems to solve a problem. Research and applications relate to semantic web and ontology provide corresponding technical basis and practical solutions to deal with the difficulties which are caused by interoperability or knowledge sharing and reuse between distributed knowledge systems. Ontology can give a clear statement of common understanding for a given domain, which will facilitate communication between all the subjects. Reusing and sharing of knowledge in a given domain shall be achieved through these exchanges.
In this paper, we studied the construction of ontology-based expert system, which include building Domain Ontology and designing expert system based on domain ontology. This work belongs to the State 863 Project "Digital Agriculture Knowledge Grid Research and Applications". And it is a part of "Web-based Distributed Agricultural Knowledge Consultation Platform".
The work in this paper:
1. Proposed a method of constructing domain ontology-"Nine-Step"
After researching the theoretical knowledge such as the concept of ontology, the main ontology description language, construction methods and development tools, this paper presents a method of constructing domain ontology-"Nine-Step". "Nine-Step" includes the following nine steps:Determine the areas and scope of the ontology; Consider reusing existing ontology; Knowledge acquisition; List terms; Define classes and levels; Defined attributes;Define the side:Define instance:Abnormal test.
2. Carried out ontology engineering under the direction of "Nine-Step", built "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology"
We manage to represent the domain knowledge of maize disease and pest diagnosis in "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology" through the establishment of the concepts and the relationship between concepts, the definition of instance and constraints over properties. We select OWL DL which is a sublanguage of OWL to describe the ontology. And we chose Protege 3.4.4 which is developed by Stanford University as our ontology development tool.
3. Using the ontology model as the knowledge base of maize disease and pest diagnostic system, we carried out the design and implementation of the system by using Visual C# 2008 and Microsoft SQL Server 2005.
We worked from the reality of maize pest control, carried out demand analysis and overall design, determined the user, functions and framework of the system. The system uses B/S structure, could be divided into knowledge layer, the data layer, logic layer and the HCI layer in accordance with the functions. System features include the diagnosis of common diseases and pests of maize, the query of the corresponding details of diseases and pest, and the management of diseases and pest. The system can not only diagnose the pests and diseases that have occurred, but also feedback the potential threat of pests to the user; system provides a variety of diagnostic mode, allowing users to choose the right one according to the actual situation.
The significance of this work:
The "Nine-Step" works well in the scene of building domain ontology by knowledge engineers, and makes ontology maintenance and evolution work much easier. Through the building of "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology", we try to construct a shared, reusable knowledge base for maize disease and pest diagnosis expert system. The ontology-based diagnosis system for maize diseases and pests is not pleasing to the eye and easy to use, but also meet user's demand for information services in agricultural production. All in all, the system has very good practicability.
本文对基于本体论的专家系统的构建进行了研究,具体工作包括构建领域本体和设计基于本体的专家系统两部分。本文的工作属于国家863项目“数字农业知识网格技术研究及应用”中“基于Web的分布式农业生产知识咨询平台”的一部分。
本文具体内容：
1.提出了一个领域本体的构建方法——“九步法”。
对本体的概念、主要的本体描述语言、构建方法及开发工具等理论知识进行了研究,提出了一个领域本体的构建方法——“九步法”。“九步法”包括以下九步：确定本体的领域和范围；考虑复用已有本体；获取知识；列举术语；定义类和层次；定义属性；定义侧面；定义实例；检查异常。
2.依照“九步法”实施本体工程,构建了玉米病害诊断本体和玉米虫害诊断本体。
在构建本体时参考了相关农学资料,并接受了农学专家的指导。通过建立概念、概念间的关系、定义实例、定义属性的约束等方式表达了玉米病虫害诊断的领域知识。该本体采用了W3C推荐的本体描述语言OWL进行描述,利用斯坦福大学开发的本体构建工具Protege3.4.4进行开发。
3.利用构建的玉米病害诊断本体和玉米虫害诊断本体作为玉米病虫害诊断系统的知识库,采用Visual C# 2008和Microsoft SQL Server 2005开发了基于“图像规则”的玉米病虫害诊断系统。
从玉米病虫害防治工作的实际情况出发,进行了系统的需求分析和设计,明确了系统的用户、功能和框架。系统采用B/S结构,按照功能划分为知识层、数据层、逻辑控制层和人机交互层。系统的功能包括对玉米常见病虫害的诊断,病虫害详细信息及对应防治方案的查询。系统不仅能对当前已经发生的病虫害进行诊断,也能将潜在的虫害威胁反馈给用户；系统还提供了多种诊断模式,支持用户根据生产中的实际情况选择合适的模式进行诊断。
本文工作的意义：
本文提出的“九步法”适用于由知识工程师构建领域本体的场景,可以较好的支持本体的维护和进化工作；通过构建玉米病害诊断本体以及玉米虫害诊断本体,为玉米病虫害诊断专家系统建立可共享、可重用的知识库提供了参考；基于本体的玉米病虫害诊断系统不但美观易用,而且满足了用户在农业生产中对信息服务的实际需求,具有很强的实用性。
Most of the current knowledge systems are developed or redeveloped base on the user's different needs, using different knowledge representation methods, different modeling methods and different development tools. Therefore, the interoperability between knowledge systems becomes a problem. not to mention coordinate these systems to solve a problem. Research and applications relate to semantic web and ontology provide corresponding technical basis and practical solutions to deal with the difficulties which are caused by interoperability or knowledge sharing and reuse between distributed knowledge systems. Ontology can give a clear statement of common understanding for a given domain, which will facilitate communication between all the subjects. Reusing and sharing of knowledge in a given domain shall be achieved through these exchanges.
In this paper, we studied the construction of ontology-based expert system, which include building Domain Ontology and designing expert system based on domain ontology. This work belongs to the State 863 Project "Digital Agriculture Knowledge Grid Research and Applications". And it is a part of "Web-based Distributed Agricultural Knowledge Consultation Platform".
The work in this paper:
1. Proposed a method of constructing domain ontology-"Nine-Step"
After researching the theoretical knowledge such as the concept of ontology, the main ontology description language, construction methods and development tools, this paper presents a method of constructing domain ontology-"Nine-Step". "Nine-Step" includes the following nine steps:Determine the areas and scope of the ontology; Consider reusing existing ontology; Knowledge acquisition; List terms; Define classes and levels; Defined attributes;Define the side:Define instance:Abnormal test.
2. Carried out ontology engineering under the direction of "Nine-Step", built "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology"
We manage to represent the domain knowledge of maize disease and pest diagnosis in "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology" through the establishment of the concepts and the relationship between concepts, the definition of instance and constraints over properties. We select OWL DL which is a sublanguage of OWL to describe the ontology. And we chose Protege 3.4.4 which is developed by Stanford University as our ontology development tool.
3. Using the ontology model as the knowledge base of maize disease and pest diagnostic system, we carried out the design and implementation of the system by using Visual C# 2008 and Microsoft SQL Server 2005.
We worked from the reality of maize pest control, carried out demand analysis and overall design, determined the user, functions and framework of the system. The system uses B/S structure, could be divided into knowledge layer, the data layer, logic layer and the HCI layer in accordance with the functions. System features include the diagnosis of common diseases and pests of maize, the query of the corresponding details of diseases and pest, and the management of diseases and pest. The system can not only diagnose the pests and diseases that have occurred, but also feedback the potential threat of pests to the user; system provides a variety of diagnostic mode, allowing users to choose the right one according to the actual situation.
The significance of this work:
The "Nine-Step" works well in the scene of building domain ontology by knowledge engineers, and makes ontology maintenance and evolution work much easier. Through the building of "Maize Disease Diagnosis Ontology" and "Maize Pest Diagnosis Ontology", we try to construct a shared, reusable knowledge base for maize disease and pest diagnosis expert system. The ontology-based diagnosis system for maize diseases and pests is not pleasing to the eye and easy to use, but also meet user's demand for information services in agricultural production. All in all, the system has very good practicability.
引文
[1]Tim Berners-Lee. Semantic Web-by XML 2000[EB/OL].2000-12-06. http://www.w3.org/2000/Talks/1206-xml2k-tbl/sild10-0.html.
[2]F Berman. From teragrid to knowledge grid [J]. Communications of the ACM.2001.44 (11):27-2.
[3]M Cannataro. D Talia. Knowledge Grid:An architecture for distributed knowledge discovery [J]. Communications of the ACM,2003,46 (1):89-93.
[4]H Zhuge. China s E-Science Knowledge Grid Environment [J]. IEEE Intelligent System, 2004,19(1):13-17.
[5]马文峰,杜小勇.知识网格研究[J].图书情报工作,2007.51(10)：64-67+135.
[6]陆静.基于知识网格的知识整合研究[J].农业图书情报学刊.2009，21(10)：19-21.
[7]闰洪森，张野,孙娜，孔德冉.杨喜权.基于本体的知识库构建方法[J].情报科学,2007,25(9)：1398-1400+1408.
[8]徐勇，高雅,刘艳华.农业过程本体及其构建方法——以玉米为例[J].农业网络信息，2009(11)：8-11.
[9]高文，张小栓，傅泽田,张领先，岳俊，李楠.基于OWL的鱼病诊断本体模型[J].计算机工程与设计，2007,28(19)：4770-4773.
[10]全国农技推广服务中心外经处.美国病虫害综合防治(IPM)概况介绍[J].植保技术与推广，1999，19(5)：35-39.
[11]邵刚，李志红等.北京地区蔬菜病虫害远程诊治专家系统VPRDES的研究[J].植物保护.2006,32(1)：51-54.
[12]闫飞燕,文仁来,张述宽,顾林.广西玉米病虫害诊断决策专家系统的建立与应用[J].广西农业科学,2006,37(4)：374-378.
[13]明博，赖军臣等.基于图像规则的玉米病害诊断系统的设计与实现[J].玉米科学.2009，17(6)：132-135.
[14]吴保国.温亮宝.森林重大病虫害诊治专家咨询系统的研究[J].北京林业大学学报,2006,28(6)：113-118.
[15]李成赞.森林病虫害诊断及害虫预报专家系统的研建[D].北京林业大学.2009.
[16]郭宏亮.玉米病虫草害诊断系统的研究与实现[D].吉林大学.2008.
[17]Gruber T R. A Translation Approach to Portable Ontology Specifications[J]. Knowledge Acquisition,1993,5:199-220.
[18]Stuber R et al.. Knowledge Engineering:Principles and Methods[J]. Data and Knowledge Engineering,1998,25(1-2):161-197.
[19]Grigoris Antoniou,Frank van Harmelen.语义网基础教程[M].北京:机械工业出版社, 2008.
[20]Natalya F.Noy,Deborah L.McGuinness.Ontology Development 101:A Guide to Creating Your First Ontology[EB/OL].2006.http://protege.stanford.edu/publicatioans/ ontology_development/ontology101-noymcguinness.html.
[21]Fernandez M.,Gomez-Perez A.,Juristo N..METHONTOLOGY:From Ontological Art Towards Ontological Engineering,AAAI-97 Spring Symposium on Ontological Engineering, 1997[C].Stanford:Stanford University,1997.
[22]Joseph Giarratano.专家系统原理与编程[M].北京：机械工业出版社.2005.
[23]刘春辰.基于本体的玉米病虫害语义检索系统的研究[D].吉林大学,2008.
[24]戴维民.语义网信息组织与技术方法[M].上海：学林出版社.2008.
[25]鲁新,张国红,李丽娟,刘宏伟,汪洋洲.吉林省亚洲玉米螟的发生规律[J].植物保护学报,2005,32(3)：241-245.
[26]Joshua Greenberg.Building a Rule Engine with SQL Server[EB/OL].2006, http://msdn.microsoft.com/en-us/library.
[27]Michael K Smith,Chris Welty.OWL Web Ontology Language Guide[EB/OL].2004. http://www.w3.org/TR/owl2guide/
[28]李朝东,崔国贤,盛畅,谢宁.李旺.农业专家系统的发展概况与展望[J].农业网络信息,2009(2)：4-12.
[29]袁福香,刘实,郭维,胡艳全.吉林省一代玉米螟发生的气象条件适宜程度等级预报[J].中国农业气象,2008，29(4)：477-480.
[2]F Berman. From teragrid to knowledge grid [J]. Communications of the ACM.2001.44 (11):27-2.
[3]M Cannataro. D Talia. Knowledge Grid:An architecture for distributed knowledge discovery [J]. Communications of the ACM,2003,46 (1):89-93.
[4]H Zhuge. China s E-Science Knowledge Grid Environment [J]. IEEE Intelligent System, 2004,19(1):13-17.
[5]马文峰,杜小勇.知识网格研究[J].图书情报工作,2007.51(10)：64-67+135.
[6]陆静.基于知识网格的知识整合研究[J].农业图书情报学刊.2009，21(10)：19-21.
[7]闰洪森，张野,孙娜，孔德冉.杨喜权.基于本体的知识库构建方法[J].情报科学,2007,25(9)：1398-1400+1408.
[8]徐勇，高雅,刘艳华.农业过程本体及其构建方法——以玉米为例[J].农业网络信息，2009(11)：8-11.
[9]高文，张小栓，傅泽田,张领先，岳俊，李楠.基于OWL的鱼病诊断本体模型[J].计算机工程与设计，2007,28(19)：4770-4773.
[10]全国农技推广服务中心外经处.美国病虫害综合防治(IPM)概况介绍[J].植保技术与推广，1999，19(5)：35-39.
[11]邵刚，李志红等.北京地区蔬菜病虫害远程诊治专家系统VPRDES的研究[J].植物保护.2006,32(1)：51-54.
[12]闫飞燕,文仁来,张述宽,顾林.广西玉米病虫害诊断决策专家系统的建立与应用[J].广西农业科学,2006,37(4)：374-378.
[13]明博，赖军臣等.基于图像规则的玉米病害诊断系统的设计与实现[J].玉米科学.2009，17(6)：132-135.
[14]吴保国.温亮宝.森林重大病虫害诊治专家咨询系统的研究[J].北京林业大学学报,2006,28(6)：113-118.
[15]李成赞.森林病虫害诊断及害虫预报专家系统的研建[D].北京林业大学.2009.
[16]郭宏亮.玉米病虫草害诊断系统的研究与实现[D].吉林大学.2008.
[17]Gruber T R. A Translation Approach to Portable Ontology Specifications[J]. Knowledge Acquisition,1993,5:199-220.
[18]Stuber R et al.. Knowledge Engineering:Principles and Methods[J]. Data and Knowledge Engineering,1998,25(1-2):161-197.
[19]Grigoris Antoniou,Frank van Harmelen.语义网基础教程[M].北京:机械工业出版社, 2008.
[20]Natalya F.Noy,Deborah L.McGuinness.Ontology Development 101:A Guide to Creating Your First Ontology[EB/OL].2006.http://protege.stanford.edu/publicatioans/ ontology_development/ontology101-noymcguinness.html.
[21]Fernandez M.,Gomez-Perez A.,Juristo N..METHONTOLOGY:From Ontological Art Towards Ontological Engineering,AAAI-97 Spring Symposium on Ontological Engineering, 1997[C].Stanford:Stanford University,1997.
[22]Joseph Giarratano.专家系统原理与编程[M].北京：机械工业出版社.2005.
[23]刘春辰.基于本体的玉米病虫害语义检索系统的研究[D].吉林大学,2008.
[24]戴维民.语义网信息组织与技术方法[M].上海：学林出版社.2008.
[25]鲁新,张国红,李丽娟,刘宏伟,汪洋洲.吉林省亚洲玉米螟的发生规律[J].植物保护学报,2005,32(3)：241-245.
[26]Joshua Greenberg.Building a Rule Engine with SQL Server[EB/OL].2006, http://msdn.microsoft.com/en-us/library.
[27]Michael K Smith,Chris Welty.OWL Web Ontology Language Guide[EB/OL].2004. http://www.w3.org/TR/owl2guide/
[28]李朝东,崔国贤,盛畅,谢宁.李旺.农业专家系统的发展概况与展望[J].农业网络信息,2009(2)：4-12.
[29]袁福香,刘实,郭维,胡艳全.吉林省一代玉米螟发生的气象条件适宜程度等级预报[J].中国农业气象,2008，29(4)：477-480.