The Design And Implementation Of RESTful Web Services Open Platform

Open service has become an important trend of the Internet, more and more enterprises and individuals release web services on the Internet. Third-party web services have become an indispensable part of our daily life, and it is an important cornerstone of the future open software. The existing Web services are mainly divided into two categories from the use of scale and quantity:the traditional structured web services represented by WSDL and the unstructured web services represented by RESTful. Because of the simplicity, light weight, good expansibility and security of RESTful web services, RESTful architecture has become the mainstream technology of web services. However, different from the traditional structured web services represented by WSDL, RESTful web services have no strict standards and rules, how to accurately and effectively identify unstructured RESTful web services has become an important research issue. Now there are various kinds of web services on the Internet, and the stability and access methods of these web services are different, how to actualize efficient integration and unified access of the identified RESTful web services is also an important research issue.In view of the above problems, this research concentrates on designing an implementation scheme of RESTful web services open platform. Platform architecture mainly consists of three modules:service discovery and identification, service search and service access. In the research of service discovery and identification module, we present a RESTful web service identification method based on cosine similarity, by defining a model about RESTful web services feature vector similarity matching, actualized the rapid integration of RESTful web services. The main steps of the method include page denoising, segmentation, removal of stop words, feature vector extraction and feature vector cosine similarity matching. Real experimental data show that this algorithm can achieve effective identification of RESTful web services, and the accuracy rate increased by more than 5% compared with traditional method. On the basis of service identification, we designed and actualized service search module. We build an index for the service library by using the full-text search engine Lucene, indexing engine will search in index database according to the search criteria that user inputs, then achieve the RESTful web services that meet user’s demands. For the complex services library, we build service access module based on Enterprise Service Bus Mule ESB to solve the deficiency of various RESTful web services in unity, stability and availability, actualize the rapid access and integration of RESTful web services.