Conceptual similarity measurement of term based on domain thesaurus

People can retrieve synonyms properly, but this tends to be loose, even in the crucially synonym-oriented systems. Domain terms can be regarded as specialized synonyms with their definitions, this paper attempts to make conceptual similarity of term measurable so as to push information retrieval and knowledge engineering in domain-specific. In the full paper, one domain thesaurus model in well-constructed conceptual classification is presented, which makes terms measurable and is universal and available in many domains; based on this model, several conceptual similarity measurement strategies are designed, which can calculate the conceptual similaritypsilas score of term; the experiment on material domain thesaurus shows that the proposed methodology can measure conceptual similarity of term effectively.