Automated query classification based web service similarity technique using machine learning

With the tremendous growth of the internet, services provided through the internet are increasing day by day. For the adaption of web service techniques, several standards like ebXML, SOAP, WSDL, UDDI, and BPEL etc. are proposed and approved by W3C. Most of the web services are operating as a query—response model. User has to submit query according to the standard adapted, and services are supporting natural language queries nowadays. The given inputs are processed by web services server can find few similarities in sentence like nouns. The keyword for nouns is filtered accurately and saved in the list as table for each domain. Same time input query words are stored in the domain. The words stored in the domain is matched with the given input queries, later used to find the similarity between the queries In this paper, an automated technique for finding web service similarity based on query classification proposed. The proposed method adapted machine learning approach called KNN, and the data maintained in a hash indexed storage tables. As a result, the relationships between the input query and stored database have been showed in precision, recall, F1-Score and Support.