Qos-based pattern recognition approach for web service discovery: Ar_wsds(Article)
...
Adarme Jaimes, Marco Antonio | 2021-08-31
Web service composition requires high levels of integration and reliability of the services
involved in its operation, which must meet specific quality criteria to ensure their proper execution
and deployment. The discovery and selection of web services currently face optimization problems.
Many services might satisfy a requirement with similar quality criteria. Because of this, software
developers have to choose the most appropriate services for a given composition, complicated
by the rapid increase in providers and services available in the cloud. Service composition also
implies coupling according to a composition flow and non-functional requirement criteria. Such requirements make selection and composition a complex task not previously solved in the literature.
This paper presents Ar_WSDS, a computational approach for web services discovery and selection
in cloud environments, which bases its implementation on the brain’s pattern recognition systematic
functioning. This process allows classifying web services through recognition modules created
dynamically based on their quality parameters, resulting in a set of web services suitable for a web
service composition. This approach allows a solution to the selection problem using less complex
tasks. This paper introduces an architectural and procedural definition that provides the web service description with a pattern to recognize and select services using different recognition levels.
We simulated our approach and evaluated it using a dataset from the QWS project that offers a set of
quality criteria collected from different providers. The web services are recognized and classified
using different quality criteria for the composition and each of their services. The results demonstrate
the effectiveness of the discovery and selection process compared to other approaches. Furthermore,
Ar_WSDS allows us to recognize and filter out web services with ambiguity and similarity in their
provider information, a process that minimizes the discovery space for services.
LEER
