| Agricultural ontology is a normative description of the conceptualization of knowledge in the agricultural field,which greatly promotes the sharing and consistent understanding of knowledge concepts in the field.The concept of agricultural knowledge is often accompanied by uncertainty,which is based on a single cloud in cloud theory to deal with this kind of uncertainty,which can not excavate the general agricultural knowledge which is more generalized,more systematic and higher,and can not embody the hierarchy structure of agricultural ontology itself.Therefore,it is very important to study how to express the knowledge of agricultural ontology from lower level to higher level.In this paper,aiming at the deficiency of the knowledge abstraction of agricultural ontology represented by a single cloud,taking the meteorological data of the tea garden of Huangshan as an example,it focuses on solving two key problems in the representation process of agricultural ontology knowledge—the atomic concept cloud extraction problem of agricultural ontology knowledge and the atomic concept cloud synthesis problem of agricultural ontology knowledge.This paper improves the deficiencies of the existing method of atomic concept cloud extraction,puts forward a new cloud synthesis criterion and a cloud synthesis method without divergence,and develops a cloud synthesizing system of agricultural ontology knowledge.The main research work of the paper is as follows:(1)Improved the concept extraction method of agricultural ontology knowledge based on backward cloud.For the defects in the backward cloud algorithm with certainty and uncertainty in the cloud theory,this paper proposes to improve the atom concept cloud extraction process of the inverse with certainty degree by Levenberg-Marquardt(LM)curve fitting,and improve the uncertainty backward cloud atomic concept extraction process by using statistical thought,and to verify the effectiveness of the improved method.(2)The method of atomic concept extraction of agricultural ontology knowledge based on cloud transformation is optimized.The optimization method using k-means++ clustering is given,which provides the optimized priori information for the Agricultural ontology knowledge atomic concept extraction,reduces the iterative times of the whole process and improves the performance of the algorithm.According to the specific examples of agricultural ontology knowledge,the appropriate atomic concept cloud extraction algorithm is chosen to complete the conversion from quantitative to qualitative concept.(3)A new cloud synthesis criterion and a cloud synthesis method without divergence are presented.Based on the traditional criterion of cloud closeness,a new comprehensive criterion is proposed,which is determined by the position of cloud center of gravity and cloud closeness degree.Aiming at the phenomenon of divergence in the process of synthesizing the agricultural ontology knowledge atomic concept cloud that has been extracted,a method for cloud synthesis of agricultural ontology knowledge without divergence is proposed,and the correctness and validity of the proposed method are validated by contrasting the experimental methods.(4)A cloud synthesis system for agricultural ontology knowledge is developed.Based on the method proposed in this research,the synthesis system of agricultural ontology knowledge cloud was developed by using python language integration under the framework of Web Flask.The system has functionally designed the data import,the agricultural ontology knowledge atom concept cloud extraction and the agricultural ontology knowledge cloud synthesis module from the function,provides the visual expression and the interactive interface for the user.In this paper,the research on the method of cloud synthesis of agricultural ontology knowledge,aiming at providing a new way for the establishment and service of modern agricultural knowledge model,which has certain theoretical research value and practical application significance. |
PDF Full Text Request