Many ontologies are provided to representing semantic sensors data.However,heterogeneity exists in different sensors which makes some service operators of Intet of Thing (IoT) difficult (such as such as semantic inferring,non-linear inverted index establishing,service composing).There is a great deal of research about sensor ontology alignment dealing with the heterogeneity between the different sensor ontologies,but fewer solutions focus on exploiting syntaxes in a sensor ontology and the patt of accessing alignments.Our solution infers alignments by extending structural subsumption algorithms to analyze syntaxes in a sensor ontology,and then combines the alignments with the SKOS model to construct the integration sensor ontology,which can be accessed via the IoT.The experiments show that the integration senor ontology in the SKOS model can be utilized via the IoT service,and the accuracy of our prototype,in average,is higher than others over the four real ontologies.