Research On Teaching-Domain Knowledge Repository Based On Ontology

As the basic components of Intelligent Tutoring System, knowledge base plays a significant role and it has been one of the important issues in the domain of intelligent tutoring system. However, the problem of traditional Intelligent Tutoring Systems'knowledge representation and organization has not been a good solution. First, descriptive standard is inconsistent that leads it is difficulty for teaching knowledge to share and interoperate between different systems. Second, the requirement of man-machine conversation, intelligent answering, auto title and paper is not being met due to lack of intelligent knowledge which computer can understand. So, we need to build a teaching-domain knowledge base to meet the characteristics and requirement of Intelligent Tutoring System and play its applied capacity.This paper built a teaching-domain knowledge base used in Intelligent Tutoring System on the core of ontology technology. The key issues involved in the build process are the knowledge representation and knowledge reasoning. The main work as follows:(1) First, on the basis of domain concepts and concept hierarchy defined in DL, the teaching-domain knowledge base utilized the defined concepts and relationships to define the new abstract concept in the TBox, so that computer can understand the intension of abstract concept. Then, the representation of production rules was used to describe causality and rules domain knowledge, so that the knowledge representation of DL could be extended. Finally, domain knowledge base was expressed congruously through Ontology language So, the teaching-domain knowledge base was established which included domain ontology and domain rules.(2) This paper discussed the reasoning mechanism of domain knowledge base according to different forms of domain knowledge. The reasoning mechanism was divided into description logic-based and SWRL-based. This one part is the abstract concept TBox reasoning through utilizing methods of description logic. It was embodied by the concept automatic classification, so that the new concept hierarchy was reasoned. The other is domain rules'reasoning based on SWRL, which extended reasoning and query capabilities of domain knowledge base through processing rules of subject knowledge.(3) This paper designed and implemented a knowledge retrieval experimental system based on the teaching-domain knowledge base. The knowledge retrieval experimental system embodied the application of the teaching-domain knowledge base in the field of ITS. The system could make students search subject knowledge more accurately and quickly. The design scheme utilized J2EE layered architecture. This paper introduced the steps and methods of system development amply.Summary, the teaching-domain knowledge base had a wealth of knowledge expression and reasoning abilities through combined Ontology and production rules. So it could support knowledge retrieval, intelligent question and answering, auto title and test paper, also meet the requirements of domain knowledge of ITS to some extent.