The Modeling And Realization Of Knowledge Of Cotton Textile Crafts And Arts Expert System

The primary problem for constructing a Cotton Textile Crafts and Arts Expert System (CTES) is how professional realm knowledge of cotton textile crafts and arts are represented and stored. This paper mainly described the studying work of describing and storing realm knowledge in object_ relation model. On the base of this the system reasoning is realized with Case-based Reasoning.Object- Oriented Knowledge Representation (OOKR) means to establish the model of the realm knowledge with the way of object-oriented standpoint and method, this is adapt to the current main object-oriented visual development environment. But if the characteristic data of the realistic business object is stored in the relation database, it is a bottleneck problem for the expert system to communicate data between the object and the relation database.Combining the object-oriented technology and relation model to represent and store the knowledge of cotton textile crafts and arts not only make it ease to establish object-oriented knowledge model and store the knowledge but also make it possible to make use of the high data independence and simple query and simple processing of the relation database, at the same time the combining is convenient for the system to communicate data between the object and the relation database.The knowledge within this system knowledge base is divided into the case knowledge and reasoning rule knowledge. This knowledge is delaminated according to the producing procedures and equipments, namely, upon a premise of a cotton product request, crafts and arts of different cotton textile equipment of different producing procedure are regard as objects. The crafts and arts of some aspect of the equipment is regarded as its attribute. It is easy to realize the system's reasoning if case knowledge is represented and stored in object-relation model. For this purpose, the case knowledge is conceptualized and formalized. The knowledge database is designed on the base of this.Case-Base Reasoning is like the way of human' s thinking and dealing with questioning. That is to say, system determines to choose an existing case and revises it to get answer of new question by compare the comparabilityof the new question case and the existing case.The realization of a Integrated CBR System Includes five circular processes, namely Retrieve, Reuse, Revise, Review and Retain. For the fuzzy matching problem of case-based reasoning of CTES, this paper presents the adjusting strategy for comparability quotiety about case selecting according to the characteristic of cotton textile crafts and arts. So a similitude evaluating knowledge database is established. The thresholding of similitude degree can be adjusted actively, namely the system can retrieve many times and the thresholding is dynamic. At first, the thresholding is small in order to convenience reuse. If appropriate case can't be retrieved, the thresholding will be increased gradually. The retrieved source crafts and arts case should be revised according to the adjusting rule in order to get appropriate new crafts and arts case, which can be also added to the crafts and arts case database. This is in favor of reuse later.It is testified by experiment that CTES prototype system can reason effectively according to the input of the users and educe appropriate crafts and arts.