| The construction of water conservancy informatization started earlier in China.With the continuous research and exploration of water conservancy workers,various software service systems have come out one after another,playing a huge role in the field of water conservancy.Restricted by cognition and technology,the software service system in the field of water conservancy has strong domain and pertinence.Data cannot be Shared between application systems,forming an information island with the boundary of specialty,department and region.The introduction of ontology provides a powerful support for an application software platform that integrates information resources,information sharing and knowledge reuse to solve the problem of information island.In this paper,aiming at the problems of high data noise and low extraction accuracy of concepts and relations between concepts in the automatic construction of water conservancy ontology,a method of automatic construction of water conservancy ontology circulation is proposed by referring to the principle of snowball movement and the concept of ontology circulation construction.The cyclic extraction technology of candidate concepts based on BP(back propagation)neural network algorithm and the hierarchical multi-corpus concept relations extraction technology based on FP(Frequent Pattern)-tree frequency set algorithm are realized,which can reduce data noise and improve the accuracy of concept and concept relations extraction.The experimental results show that the automatic construction method of water conservancy ontology cycle designed in this paper is feasible.This method mainly consists of four steps: 1.Collect the required unstructured data as learning data,and then conduct data preprocessing;2.Secondly,BP neural network algorithm is used to extract the concepts of water conservancy;3.The fp-tree-based frequency set algorithm is used to extract the relationship between concepts;4.Evaluate the generated domain ontology. |
PDF Full Text Request