The Research And Application Of Water Environmental Ontology Construction And Organization Methods

With the development of water information technology, because there is no unified form of information expression, it has caused that a lot of information can't be shared and integrated, which restricts the progress of water science and technology. Ontology can be very good application as a standard form of expression. In order to share the information about water environment, we need to construct a domain ontology in the field of water environment as the research object. In this paper, water environment domain ontology construction and organization methods are proposed, and construct the water environment ontology, which contains concept and the relationship between the concept in the field of water environment.Considering the limitation of existing methods about ontology construction, we put forward a way by combining seed concepts method and TFIDF method to extract concepts. Taking into account the uneven distribution of the words in the text, the formula of original TFIDF has been improved, which improves the accuracy of the ontology concept extraction. What is more, many seed concepts has been selected from Semantic Web for Earth and Environmental Terminology Ontology and CUAHSI Ontology, which are top-level ontology that contain many concepts about water environment. The approach is integrated applied to alleviate the drawbacks of TFIDF method, such as less extraction of low-frequency words. Then, the candidate concept is filtered by domain relevance and domain consensus. Secondly, a K-means-GA hybrid algorithm to identify classification relations between extracted concepts is proposed. The hybrid clustering method improves the precision of the clustering analysis by genetic algorithm, which has a global optimization search ability to overcome the randomness of the artificial selection of the K value and the limitation of the generated ontology hierarchy. Again, multi-strategy ontology mapping by calculating the similarity of two concepts are presented, which is consist of concept name, concept structure, concept attribute and concept examples. Sigmoid function to determine the weight of each strategy is introduced, which can improve the efficiency and accuracy of ontology mapping. Water environment ontology organization has been finished.Finally, a water environmental ontology construction and organization system is designed and developed. According to the ontology construction method and process, the system is divided into five modules: pre-processing module, concept extraction module, the relationship of concept extraction module, ontology mapping and integration module, ontology visualization module. Ontology construction and visualization from text data source can be completed. In experiment, the validity of the ontology construction method is verified, and the technical support for ontology construction and ontology application is provided.