Research On Service Clustering Algorithm Based On Semantics In Internet Of Things

Clustering IOT services can effectively improve the efficiency of IOT service discovery.However,in most of the existing IOT service clustering methods,their service description language is WSDL language and clustering algorithms are traditional clustering algorithms.The WSDL language cannot semantically describe IOT services,which is easy to cause misunderstanding,while a traditional clustering algorithm may have some defects,such as an over-dependence on the initial clustering center or a weak global search ability.These problems affect the accuracy of clustering results.Therefore,this thesis studies the service clustering algorithm based on semantics in Internet of things.The specific research contents of this thesis are as follows:(1)Describe IOT services consistently through semantic processing,which helps to avoid the misunderstanding caused by different description methods.Choose the appropriate semantic service description language OWL-S language,and then analyze the existing semantic transformation methods of IOT services.In view of the shortcomings of these methods,this thesis proposes a new transformation method from WSDL file to OWL-S file.In the new method,the transformation from WSDL file to OWL-S file is realized by operation mapping and ontology mapping.(2)The K-Means algorithm has been widely used in the clustering of IOT services,however,it is generally applied to the clustering of IOT services directly and the defects of the algorithm are ignored.This thesis presents a genetic K-Means clustering method combined with simulated annealing algorithm,and this new algorithm helps improve the performance of the K-Means clustering algorithm and optimize the clustering results.(3)Convert the WSDL service files in the test set into OWL-S service files based on the transformation method proposed above,and then formalize the OWL-S files into semantic vectors through a vector space model extended by semantic similarity to provide data for the service clustering algorithm.Using the K-Means algorithm and the improved algorithm to cluster the WSDL service files and then do the same for the OWL-S service files.The experimental results show that the accuracy of IOT service clustering results has been significantly improved after the semantic processing of IOT services and the improvement of the clustering algorithm.