An Ontology-based Semantic Item Bank&Knowledge Mining Research And Application

Traditional item bank system is mainly stored in text format, which lacking ofsemantic information, and the computer can not understand the specific meaning of thequestions in exercises; Secondly, there are variety of performance in item bank, and theitem bank resources are difficult to share and reuse. Moreover, traditional item bankonly summarized the knowledge of exercises, while ignoring the problem-solvingmethods of exercises. How to design a set of semantic features, and the computer canunderstand, and the item bank resources can mutually share and reuse has become animportant subject of research and development of knowledge engineering.In order to make the resources of item bank effectively share and reuse, and thecomputer can understand the questions in exercises expressed, we use the ontologytechnology into item bank system, and put forward a method to build the semantic itembank system. This method is that creating the knowledge ontology based on semanticitem bank by mining the impliedly knowledge information and the problem-solvingskills from exercises. And this thesis focuses on the knowledge ontology based onsemantic item bank and knowledge mining in-depth studying. The main research resultsare as follows:First, we associate the exercise with ontology to make the item bank has primarysemantic information by building the domain ontology. Furthermore, we in-depthknowledge of mining for the exercises, and use the makeup elements to pretreat theexercise into XML so as to make the item bank more semantic function. We will extractthe knowledge and problem-solving approach to handling the abstract model andinstance model, and then creat the process ontology by using the abstract models andinstance models. Thus, the item bank process ontology makes knowledge and theproblem-solving methods in the exercise full of sharing and reuse.Second, this thesis applied to the middle school mathematics by using the methodof the construction of semantic item bank. Furthermore, we design and implement apersonalized intelligent educational tutorship system based on the semantic item bank.The system tutors the students by using the interactive learing modes, and records the student’s learning data. And then, the system provides personalized intelligent tutoringservice through each student’s learning actuality. It successfully applied to the automaticproof through the semantic item bank and fully proved that the ability of practicalapplication.The experiments show that the semantic item bank system improve the utilizationof the item bank resources. Knowledge and information in the item bank has get theeffective sharing and reuse. Therefore, the semantic item bank provide a new idea forthe research and the application of the item bank system. Similarly, it is significant forthe research of the computer-aided education.