LWSDM: Layered Web Service Discovery Mechanism

The boom of Service oriented architecture is on substantial rise. SOA developers orchestrate individual SOA objects into web services. It has been a dream of software engineers to promote and innovate the concept of software reusability. Providing and using services dynamically over web is increasingly becoming times demand towards promoting software reuse in real sense. Probably the most important aspect in dynamic web service access is the web service discovery. Many approaches and frameworks have been proposed to discover web services. Some of the approaches assume that the requests are placed in SOAP compatible formats whereas some works are based on handling the plain text queries by their semantics while others focus on key word based query processing. The challenge remains to address the issue of handling request queries and presenting a set of suitable services to the service requester. We use the principle of compositionality to analyze and parse the request query. We have used a lightweight NLP approach by parsing the request query into constituent parts of speech wherein the verbs are used to derive the interface names and the set of identified interface names are analyzed against the nouns and adjectives for their appropriateness. This paper tries to propose a Layered architecture that the organizes the method of web service discovery in an efficient and structured manner using a intermediary called the Request Analyzer between the service provider and service requester via a service broker. We describe how the RA will be facilitating the processing of a plain text request query to finally being resolved to a most appropriate web service. We propose an algorithm for a complete cycle of web-service discovery. We describe a cache based service broker approach for less time-consuming service discovery path. We also incorporate a mechanism for keeping track of the trustworthiness of a web service being used and hence ensuring the authenticity for future use.