Semantic Matching Algorithm Design And System Implementation For Semantic Web Of Things

To solve the inherent contradiction in Internet of Things, Semantic Collaboration is introduced, so that Semantic Web of Things is formed as a new generation network. But two main problems still exist in Semantic Collaboration. First, semantic mismatch is not conductive to the global sharing of things’information; second, expressing skill lackage of traditional logic and semi-decidability of final system cause the unsustainability of reasoning (logic formula computability).To solve above-mentioned problems, a logic formula deduction based semantic matching algorithm suitable for Semantic Web of Things is proposed, which makes up defects that most existing structural similarity algorithm have. In this algorithm, the improved matching algorithm of S-Match is used to reason at conceptual level in knowledge base, and CL4 of computability logic is employed to reason at instance level in knowledge base. And, by transforming semantic coordination from structural similarity computing to logic formula deduction, automaticity and integrity of reasoning can be ensured so that reasoning sustainability can always be kept.Based on above-mentioned algorithm, a novel semantic matching framework for Semantic Web of Things is designed. This framework is composed of fives functional modules such as ontology loading, knowledge base consistency maintaining, inference rule construction, ontology reasoning and services recommending. Meanwhile, a semantic matching prototyped system for Semantic Web of Things is analyzed and designed, and techniques of natural language processing and visual programming are utilized in its implementation. Finally, a benchmark test set is employed to examine precision ratio and recall ratio of the proposed algorithm and system. The experimental results reveal that the proposed algorithm and designed system in this paper can effectively provide users with high-quality semantic collaboration service.