| Ontology outlines the basic knowledge system of a certain field through the standardized description of concepts,terms and their relationships.It can solve the problems of knowledge representation,knowledge organization and knowledge sharing well.On the basis of XML and RDF,ontology layer is responsible for describing the concepts and relationships between concepts in related fields,which provides the basis for logical reasoning and functional verification of semantic web.At present,the research of ontology is mainly based on domain ontology.The ontology that has been constructed is mostly aimed at a certain domain,or even subdivision domain knowledge.With the development of science and technology,the intersection between disciplines is more and more,especially the intersection between subdivision fields within disciplines is more frequent.It is time-consuming and laborious to develop an ontology that integrates the knowledge of the corresponding domain for each requirement,and the resource cost is large and the knowledge reuse rate is low.Aiming at the problem of agricultural multi-ontology knowledge sharing and reuse,this paper introduced virtualization thought,constructed an agricultural ontology cloud,integrated the agricultural ontology in network environment,acquired relevant ontology knowledge on demand,and realized the dynamic allocation of agricultural ontology resources in the network by integrating the dynamic scalable virtual ontology.The main research work and results are as follows:(1)On the basis of the original ontology in the agricultural field,the agricultural ontology cloud is extracted according to the need to generate the agricultural virtual ontology.In essence,the agricultural ontology cloud is composed of several ontology knowledge modules related to demand,which are physically separated and logically connected.In the final analysis,it is distributed in each agricultural domain ontology.Therefore,the construction of agricultural ontology was the premise of forming agricultural ontology cloud.Firstly,this paper gave the construction steps of tea ontology,and then collated the concepts of tea disease,tea pathogen and tea pest through the relevant literature in tea field,and formed the corresponding conceptual structure.And then use Protégé to create concepts,relationships and axioms to formalize the formation of tea diseases,tea pathogens and tea pest ontology.Finally,the Fa CT++ inference machine was used to infer the constructed ontology,which had no inconsistency,and the effect of ontology evaluation was good,which provided research and experimental objects for agricultural ontology cloud modeling.(2)In the development of agricultural ontology services,it often does not need all the knowledge of a agricultural ontology to support the required services,but only a part of the knowledge in the ontology can complete the ontology services.Even when multiple ontologies are needed to carry out collaborative services in a network environment,only part of the knowledge of these ontologies is often needed,so when constructing the agricultural ontology cloud.It is necessary to extract some knowledge modules related to requirements from several related ontologies to form a set of knowledge on demand.Therefore,this paper proposed a method of agricultural ontology knowledge module extraction based on community division and local judgment.Firstly,the network properties of tea disease,tea pathogen and tea pest ontology were analyzed.Through statistical analysis of degree,degree distribution and clustering coefficient of concept nodes,it was concluded that the network structure of the three ontologies was non-standard of complex network,and also had the characteristics of small world.Then the method of network community division was introduced into the ontology knowledge module extraction,and then the ontology was divided according to the demand,and then the ontology with community structure was judged locally.The ontology knowledge with non-local relationship with demand was obtained to realize the extraction of agricultural ontology knowledge module.The experiment showed that the ontology can be divided into communities before judging the localization of ontology knowledge,which can effectively improve the correlation between the extracted ontology knowledge module and the demand,and improve the extraction quality of ontology knowledge module.(3)In the development of agricultural ontology services,it often does not need all the knowledge of a agricultural ontology to support the required services,but only a part of the knowledge in the ontology can complete the ontology services.Even when multiple ontologies are needed to carry out collaborative services in a network environment,only part of the knowledge of these ontologies is often needed,so when constructing the agricultural ontology cloud,It is necessary to extract some knowledge modules related to requirements from several related ontologies to form a set of knowledge on demand.Therefore,this paper proposes a mapping method of agricultural ontology knowledge module based on synonym lexical forest.Firstly,the structure of synonym lexical forest and the characteristics of tea knowledge were analyzed,and the existing synonym lexical forest was extended accordingly.Then,the semantic similarity algorithm matching synonym lexical forest was given.According to the extended synonym lexical forest structure,the related parameters were determined and the mapping of agricultural knowledge module was realized by calculating the conceptual similarity between ontology knowledge modules.The experiment showed that through the extended synonym forest,the mapping of agricultural ontology knowledge module can be carried out,which can close to the knowledge characteristics of agricultural field and improve the mapping quality.(4)Several ontology knowledge modules extracted according to demand belong to several agricultural subdivision ontology,but the purpose and method of ontology construction in each agricultural field are different.The structure and expression methods of ontology knowledge modules extracted from them are also different,and they can not be inferred directly from each other.Therefore,this paper constructed a virtual ontology generation model based on HBase and RBAC in agricultural field.The ontology storage model of agricultural field was constructed based on HBase,and the RBAC was introduced into the storage model.According to the extraction and mapping of agricultural ontology knowledge module,the method of adding role to HBase line key was adopted.Control the access range of ontology knowledge and generated agricultural virtual ontology.Experiments showed that the larger the scale of ontology,the higher the efficiency of the model to generate agricultural virtual ontology.(5)The agricultural virtual ontology is still composed of ontology modules scattered in various places,and each ontology module has neither physically merged on the semantic web ontology layer,nor extracted from its own ontology.This makes it necessary to combine the distributed method for logical detection of agricultural virtual ontology.Therefore,this paper proposes a method of agricultural virtual ontology logic detection based on distributed Tableau.Based on ALC,the description logic is extended to SHOIN(D)corresponding to OWL.Then,the description logic of virtual ontology that introduced ideas into is given,to form the method of distributed logic detection oriented to virtual ontology.The feasibility of the method is proved by an example.The research results obtained in this paper can further improve the efficiency of knowledge sharing and reuse of agricultural ontology,reduced the complexity of ontology collaborative reasoning,improved the efficiency of ontology service,and realized the acquisition of multi-ontology knowledge on demand.It has certain theoretical value and practical significance to promote the development of ontology-based agricultural knowledge. |
PDF Full Text Request