Research On Key Technologies Of IA Oriented Case Knowledge Management System

The 21st century is a knowledge-economy era. The management of knowledge is the thrust of the development of our social economy and the important means of improving the competitiveness of an organization. Case Knowledge, as a special category of knowledge, is rich of information and knowledge. The main form of case knowledge management is CBR (Case-Based Reasoning) nowadays; Using CBR to manage knowledge has many advantages, but now the appeared theory models of CBR have some flaws, and the CBR systems lack of universal, both of them are Constrain the developing of CBR. The takeoff of traditional thinking of AI (Artificial Intelligence) is trying to understand the cases and establish models of them, and provide users with solutions when in use. But the takeoff of new thinking of IA (Intelligence Assistant), which was appeared in recent years, is trying to describe the nature of the cases and their relations, and help users to solve problem when necessary, artificial intelligence is just an auxiliary assistant here. A new reasoning technique will come into being by combining CBR with IA, CBR can be applied to any field ,even do case-based reasoning on the basis of no case.In this paper, the existing theoretical models of CBR are analyzed firstly, and a new theoretical model named CBR~2 model, which is proposed by applying IA-oriented artificial intelligence technology to the process of CBR by our research team, is addressed and analyzed in details. Then, some key technologies of CBR~2 model oriented case knowledge management system are researched. And a framework of case knowledge management system is analyzed and a prototype system is designed and implemented at last.The contribution of this paper is mainly reflected in the following aspects:1) Has improved and perfect the new CBR model——CBR~2 model.2) Participated in the research of SVSM (Sentences-based Vector Space Model), and implemented the retrieval algorithms and classification algorithm of case based on this model and obtained a good retrieval effect.3) Researched the index system of case based on SVSM, and by which, we improve the efficiency of the retrieval algorithms a lot.4) Designed the system architecture of case knowledge management system according to the thinking of IA and CBR~2 model.5) Researched the library technology of dynamic methods by JAVA programming language, which makes the retrieval algorithms can be dynamic extended, and provides to software evolution and case retrieval by retrieval algorithm of special domains.6) Finished the code of the overall framework and some modules, such as case base building, case resources management, textual case classification and retrieval, case attributes extraction, digital case reasoning, and domain knowledge management, of a case knowledge management system by using the advancedAOM framework.