鈦酸鍶鋇(ba0.7sr0.3tio3)熱釋電陶瓷材料研究

1、河南科技大學(xué)碩士學(xué)位論文基于相似度計(jì)算的本體映射的優(yōu)化姓名：谷志鋒申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：計(jì)算機(jī)應(yīng)用技術(shù)指導(dǎo)教師：劉勇;郭跟成20080601摘要 IISubject: The improvement of Ontology mapping based on the computation of similarity Specialty: Computer Applivations Technology Name

2、: Gu Zhi-Feng Supervisor: Liu Yong Guo Gen-Cheng ABSTRACT Along with the application of ontologys in many domains, the integration of different ontologys became an import

3、ant problem of the research of Ontologys. Now the heterogeneous ontologys ’ integration was the most difficult problem in Ontology’ research. The ontologys ’ heterogeneity included semantic heterogeneity、structure hetero

4、geneity etc. Ontology mapping was the best method to solve the integration of the heterogeneous ontologys. There are many methods to solve the ontology mapping, such as Schema mapping、the mapping based on super Ontology

5、etc. The mapping method based the computation of similarity became the popularest method for the manageable of the computation of similarity. But there were some shortages in this method, such as the vast runtime and com

6、putational unilateralism of the computation of similarity in Ontology mapping. This paper was to meliorate the shortage, then improve the performance of Ontology mapping. Firstly, the paper used filtration method to solv

7、e the vast runtime of the computation of similarity. The paper used method of candidate mapping to reduce the amount of concepts being going to mapping. And the paper used method of information gain to reduce the amount

8、of attribute being going to compute. Consquently to filtrate the insignificant concepts and attribute and reduce the runtime of the computation of similarity. In the last , experiment indicated that the effect of those c

9、oncepts and attribute being got rid of was very little. Secondly, the paper used synthetical computation to solve the computational unilateralism of the computation of similarity. The paper synthetized the concept name s

10、imilarity and instance similarity to get the basic similarity. Then the paper used the basic similarity to compute and optimize the attribute similarity、structural similarity and relational similarity. Synthetized these

11、results to get a more consummate result. At the same time, the paper used rational axiom in the computation to make the result more perfect. In the end, the paper designed experiment to prove the method in the paper. Exp