Research On The Key Technologies Of Case Knowledge Base Based On Sentence Vector Space Model

With the development of information technology, empirical knowledge management has been a core competitive power of large enterprises. Unstructured empirical knowledge accounts for a large proportion of human empirical knowledge and case knowledge is an important part of unstructured empirical knowledge. At the same time,efficient exchange and efficient sharing of case knowledge has been a source of solving new problem with empirical knowledge. The case Knowledge base, which store many cases, is the foundation and necessary condition for Case-based reasoning, therefore, research on case knowledge base is significance.This paper mainly studies the key technologies of Sentence Vector Space Model based Case knowledge base and the main works are as follows:1. A case based reasoning cubical model(CBRCM) is propos led in this paper and it mainly includes case entity layer, case abstract layer and domain knowledge layer. Case entity layer mainly supports the traditional case based reasoning. Case abstract layer mainly abstracts problems and solutions from cases. Domain knowledge mainly explains and supports cases and case-based reasoning with relations among semantic, domain knowledge and concepts to improve effective of case knowledge management and case knowledge sharing.2. The case based reasoning cubical model is just a structural model. In order to apply this model in the actual development of case knowledge management, the mainly parts of case knowledge base, such as case knowledge, taxonomies, Concepts, resources, domain knowledge, rules and so on are modeled with ontology and this will enhance the sharing, reuse and interoperability of case knowledge.3. A Sentence Vector Space Model Based case knowledge representation method is propos led in this paper, which upgrades the granularity of case knowledge representation from words or terms to sentences. Therefore, it supports case knowledge retrieve based on semantic and domain knowledge.4. This paper realizes case knowledge automatic clustering with text clustering to abstract case knowledge to improve case knowledge retrieve.5. A case knowledge base system based on ontology and sentence vector space model is designed and realized to check the core academic ideas of this paper and then the design idea, the development technologies, organization, workflow and functional modules of the system are introduced.