An Ontology-based Geometrical Information Retrieval System Research And Application

With the rapid development of information technology, international technology has a great influence on traditional field. A new method to share common information and learn quickly by international technology as well as information science technology comes to the world is called E-learning, which is also named online education or online learning. As the name implies, online learning is a new teaching mode bases on the Internet. It breaks the traditional time of study and limitation of space. As long as the Internet exits, users can study wherever they are.However, online learning mode is still in its infancy. Most knowledge database of this kind of system just moves course into Internet in a simple way. Spoon – feeding is the main teaching mode, and the common search engine doesn’t work well in terms of semantic retrieval of knowledge database. In order to solve these questions, this paper proposes a method using a XML attribution labeling to transform traditional exercises into structural text involving preliminary semantic information and designs a integrated coherence degree arithmetic to solve the question of semantic query on the basis of ontology. The main contents of this paper are:First, we have done a basic study on ontology and research engine. It consists of the description of ontology in a formalized way、the construction of ontology 、the category of research engine and the defect of research engine. The key point of this study is how to construct the ontology of Geometry in junior high school.Secondly, according to the study of research engine principle and the combination of ontology technology, we propose a comprehensive similarity algorithm, which transforms the calculation of similarity into the overall measurement of four influencing factors. They are morphological distance、semantic distance 、semantic overlap and concept hierarchy difference. Using analytic hierarchy process, we get the weighted value of these four indexes. In the end, we design a comparative trial in which we test the four different similarity algorithms on two ontology. Besides, we also analysis the experimental results to prove the practicability of this algorithm.Thirdly, we design a message retrieval algorithm on account of ontology on the basis of similarity comparative algorithm. This algorithm divides the whole retrieving procedure into three departments: based on the extension of concept set, the acquisitionof candidate concept and the calculation of comprehensive similarity. Via the similarity calculation based on ontology, we build a relationship between the information users have checked and semantic knowledge database and a detailed process of searching algorithm.Finally, we design a system of geometrical message retrieval. By achieving it and combining the systematic test, we affirm the application value of it. At the same time, it provides a new direction of search in the geometric field in the future.