Acquirement of class relations for ontology based on language expression and knowledge structure

Ontology plays an important role in information retrieval and organization, but it requires huge human labor in artificial construction, in particular it is even more labor-intensive and time intensive when determining concept class relation. The paper have put forward three methods to acquire class relations by machines, the first is based on terms co-occurrence, the second is based on term truncation, and the third is based on the meaning of terms. With these three methods, it is essential to acquire equivalent relations, sub-classes relations, and relevant relations by machine. Via artificial interference, these relations can be found in ontology or thesaurus. This work is of realistic significance in the construction of ontology and thesaurus and can save large quantities of labor and time.