Enhancing web service selection using enhanced filtering model
The upward trends in web service providers, consumers as well as web services pose remarkable challenges in the area of web service description, discovery and selection. While remarkable works have been done in web service discovery, selection still remains an area of challenge. Therefore, emphasis...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/37083/5/AjaoTajudeenAdeyemiMFSKSM2013.pdf http://eprints.utm.my/id/eprint/37083/ http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:69190?site_name=Restricted Repository |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The upward trends in web service providers, consumers as well as web services pose remarkable challenges in the area of web service description, discovery and selection. While remarkable works have been done in web service discovery, selection still remains an area of challenge. Therefore, emphasis is being placed on how to find an optimal service that satisfies requester’s functional and non-functional requirements. Majority of the existing approaches either ignore the role of user's non-functional requirements or place unnecessary burden on the requester to provide weights for QoS parameters having specified QoS constraints while others assigned arbitrary value of zero to the weight(s) of parameter(s) not specify in the constraints by requesters. All these have the tendency of generating bias results. This research work proposes an enhanced method for selecting optimal service for requesters using Enhanced QoS-based Web Service Filtering Model. The approach of this work differs from the previous approaches in that user’s preferences are taken into count, and the weights are derived from the constraints specified by the user. The methodology used involves exploiting requester’s specified QoS constraints to remove those services that failed in meeting those constraints from the list of services that match his functional requirement. The QoS of the filtered services are normalized using min-max method. The QoS score for each service is computed, and finally, the services are ranked in order of their QoS performance. The service with the highest QoS performance is then returned as best service to the requester. Experiments are conducted using Quality of Web Services datasets and the results confirm the model’s ability for selecting best web service based on requester’s preferences while out performing previous approaches. The outcome of this research could be adopted for solving service-oriented selection problems |
|---|