Research on ontology construction and information extraction technology based on wordnet

Introducing ontology in the field of information extraction can effectively improve the performance of information extraction. Ontology of college teachers' resumes are constructed in the form three layers of ontology framework structure on the basis of exploring the related technologies and criterion for building ontology, the Racer inference engine is used for realizing the consistency and accuracy detection. WordNet similarity calculation method is improved on the basis of this, so that manual collection method can be integrated into WordNet semantic similarity, thus substantially improving this method. The test shows that the result precisions based on WordNet ontology construction and information extraction both see significant improvement, which sufficiently shows the feasibility and validity of the method.