Internet of things service processing method based on improved k-means algorithm and semantic information

Abstract

In the context of the increasing number of Internet of Things services, it is needed to match and select Internet of Things services. However, current methods are prone to misunderstandings. Based on the goal of improving the efficiency of Internet of Things service discovery, this study proposes a new semantic transformation method for Internet of Things services using the Service Ontology Description Language. To cluster the transformed files, an improved K-means clustering algorithm based on genetic simulated annealing algorithm is proposed, and the transformed files are vectorized. The results demonstrated that under the K-means algorithm, the maximum and minimum clustering accuracy of the converted files were 87.78% and 74.44%, respectively. On the Iris dataset, the accuracy of improved K-means, the average values of the Davidson burg index, and the average contour coefficient were 94.21%, 0.4831, and 0.9533, respectively. The differences in clustering accuracy and time consumption before and after improving the K-means clustering algorithm were statistically significant (P < 0.05). The semantic transformation method for Internet of Things services designed in this article and the improved K-means clustering algorithm both had good performance. The research results can provide support for the processing of Internet of Things services. The innovation of the research lies in the design of a new semantic transformation method for Internet of Things services and combines genetic simulated annealing algorithm and K-means clustering algorithm. The research is limited by the fact that improving the algorithm involves a multitude of parameters, the determination of which would require a considerable investment of time and resources.