Research On Knowledge Discovery And Reuse Based On Rubber Material Library

As one of the mainstream materials in the world,rubber has been widely used in transportation,industry,aquatic products and military.With the rapid development of society,more and more data about rubber materials are increasing exponentially.On the one hand,these data may not be reliable and practical for us,but at the same time the important information hidden behind the data has not been excavated.On the other hand,with the increasing of rubber products,a large amount of design knowledge has been accumulated.In order to solve the problem of finding valuable information from mass data and to efficiently manage and reuse rubber product design knowledge,the paper proposes a method of knowledge discovery and reuse based on rubber material data and establishes a knowledge base management system of rubber material.Firstly,the related theories of knowledge discovery and knowledge reuse are analyzed and collated by consulting the domestic and foreign data,which lays the foundation for the later research.Secondly,the research method of knowledge discovery based on rubber text is expounded.Through establishing the knowledge discovery process model of rubber material text,the statistical analysis method is introduced in detail from two aspects:feature extraction and similarity calculation.The experimental process and results are also presented.The feature extraction is based on the x~2 statistical method,which obtains the feature words by word segmentation,preprocessing,word frequency statistics and feature selection,while the similarity calculation is based on the selection of feature vectors.The results are obtained in the process of determining the weights of feature items based on TF-IDF algorithm and calculating the similarity based on cosine coefficients.Then,the research method of knowledge reuse based on CBR is introduced.The reasoning process includes case representation,case retrieval,case modification and case storage.Case retrieval adopts combination weighting method of AHP and CVM,and case modification adopts artificial intervention and case-based modification.Case storage mainly selects whether the new instance is stored according to the relative relationship between cases.Finally,a fully functional knowledge base management system of rubber materials is developed by using the B/S mode,Tomcat server,JSP technology and MySQL database in,which provides technical guidance and reference for knowledge acquisition and product development.The results show that the proposed knowledge discovery method based on feature extraction and similarity computing can not only extract feature words from rubber text,but also calculate the similarity between two rubber texts.The knowledge reuse method based on CBR technology realizes the rapid design of rubber products and improves the efficiency of product development to a great extent.