Semantic similarity analysis between high-level model description text and low-level implementation text for network survivability

The transformation from high-level model description to low-level implementation for network survivability may lead to semantic inconsistency problem, since the method of transformation is based on symbolic transformation of machine which neglects semantics of proposed problem. Therefore, in order to analyze the semantic differences of transformation before and after and ensure the semantic consistency, this paper built an Ontology model of high-level model description and low-level implementation for network survivability. We proposed an Ontology-based multi-factor synthesized method which is used to analyze semantic similarity of multi-hierarchy text including high-level model description text and low-level implementation text. This method included the computation of type similarity of the text based on concept lattice, computation of content similarity of the text based on ontology and computation of structure similarity of the text based on inverse order pairs. The availability of the method is verified by experiments and its accuracy is proved by comparison with classic VSM algorithm, WeiSong algorithm, CF algorithm and professional value.