Abstract:

In order to solve the problems that extraction efficiency and the accuracy of Chinese domain ontology concept based on unstructured text is not ideal. We present a method of domain ontology concept extraction based on semantic rules and association rules. A set of candidate concepts and relationships are obtained by using field consistency, correlative checks and association rules, and the depth and breadth of relations of every concept in candidate concepts are computed, using the depth and breadth information to feedback the degree of membership between terminology and field, with the way of quantitative analysis to complete the extraction of domain ontology concepts. The experimental results show that this method has feasibility and rationality, the concept of domain ontology extraction accuracy increased by about 12%.

Key words: ontology concept extraction, association rules, semantic rules, domain membership checking