The Discovery Of Unnamed Concept And Its Application In Analogical Reasoning

With the rapid development and application of the internet technology,as well as the explosion of information technology,the information data which people need to deal is also increasing daily.In information data,these contained unnamed concept and named concept which have been recognized by human beings.How to retrieve the information of the unnamed concept from the massive information data,and use unnamed concept to get more complete knowledge has become the problem,that many scholars studied always.In the process of analogical reasoning,researchers found that there are some problems in the application of analogical reasoning:(1)In the traditional analogical reasoning method,the knowledge of the source domain only contains the identified concepts.And it usually analogical reasoning to get the knowledge of the goal domain based on the identified concepts.However,during the automatic learning of the knowledge,large-scale data often exist many unrecognized "quasi-concept"-the unnamed concept(the connotation and denotation can be calculated,but can't been named by humans).(2)In the existing analogical reasoning methods,there is a lack of use of unnamed concepts.With the increase of the knowledge of the source domain,knowledge management and retrieval overhead will rise rapidly.In order to find a more complete result of analogical reasoning,it is urgent to establish a proper way to use the unnamed concept.In view of the above two problems in the process of analogy reasoning,this paper will focus on how to find the unnamed concept,and use the unnamed concept analogy to get the knowledge mapping between the source domain and the target domain.The main research contents are as follows:(1)We use the concept lattice model to establish the concept relations of the source domain and the target domain.For the constructed concept lattice model,firstly,matches each of the attributes in the domain and the concepts in the ontology.And then add concepts that are not able to match the concepts in the ontology to the unnamed concepts set.Secondly,calculate the semantic value of the two concepts.If the calculated semantic value is lower than a specific threshold specified by the domain experts,then the concept of the domain is added to the unnamed conceptual set.(2)In the existing analogical reasoning model,the unnamed concept cannot be utilized very well.To focus this problem,we observed that the information is fully expressed by the concept lattice model,and can express all the concept information one by one.This paper proposes an analogical reasoning method by unnamed concept on the basis of concept lattice.Firstly we use the concept lattice model to establish the concept relations of the source domain and the target domain,and organize the data of the source domain and the target domain by the concept lattice structure,and the connection between the data(property)is expressed by the relationship between the generalizations of the concept lattice nodes.Secondly,we acquire the nodes mapping relations between the source and the target by exploit the isomorphic algorithm.Finally,we utilize those unnamed concepts to make analogical reasoning from source domain to target domain by structure mapping relation,so as to improve the effect of analogical reasoning.