Construction Of Ontology For Automatic Problem Solving Of Geography

Automatic problem solving as an important area of artificial intelligence has also attracted the attention of domestic and foreign experts and scholars.However,there are still some limitations of automatic problem-solving.Firstly,automatic problem solving has been greatly developed in mathematics,but others not.Secondly,the traditional database can not describe the relationship between the concepts of the entities,and also it is very uncomplicated.In order to solve the problem of knowledge base,we use the ontology as the knowledge base for problem-solving,because of the conceptual model.But in practice,we found that if the manual construction of the ontology knowledge base,it will cost a lot of labor.(1)For the problem proposed,we select geography as the discipline for the automatic problem-solving.Firstly we use the TF-IDF algorithm to extract the concepts of the entities with the question text.Then we design the basic term hierarchical relation table to realize the extraction of the relationship between the concepts of the question text.Finally,we use the Jena reasoning machine to further rule the relationship between entity concepts into OWL ontology and realize ontology construction and reduce the labor costs.(2)We use the ontology that we built and combine with the test text,and write relevant rules to realize and visualize the automatic problem-solving system.Design related experiments,the results show that the ontology construction method,proposed in this thesis,can construct the ontology knowledge base effectively and reduce the labor costs in a considerable degree.It has significant effect when using the constructed ontology to the automatic questions-solving system.