Ranking Machine for Web Service Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current web services architecture, specifically UDDI/WSDL, lacks the ability to incorporate originator-specific preferences for service selection, leading to generic service choices that do not account for individual user preferences such as cost, efficiency, or reliability.
Innovation Solution
A method and system that introduces a ranking machine with a choice algorithm based on the originating service requestor's preferences, allowing for biased service selection within the hierarchy of web services, where the ranking machine applies the algorithm to sets of possible lower-level services, ensuring that the selected services align with the user's preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a directory or registry is used to assemble software in a componentized fashion with generic service selection criteria, then service selection is open and accessible to all users, but the service selection cannot account for individual user preferences such as quality of service, cost, efficiency, or reliability
Solution Approach 1:
The service selection mechanism is segmented into multiple independent components: a directory/registry for service listing, a ranking machine for preference-based filtering, and a service selector for final choice. This allows the system to maintain open access while incorporating personalized preferences without creating a monolithic complex system.
Solution Approach 2:
A ranking machine is introduced as an intermediary component between the directory and the service selector. This intermediary receives service listings from the directory, applies user-specific preference criteria (quality of service, cost, efficiency, reliability), and returns ranked results to the selector, thereby enabling personalized service selection without direct user intervention in the selection process.
2Ease of operation
If service selection is made generic and open to all users through a public directory, then accessibility and openness are improved, but the ability to apply originator-specific criteria such as negotiated preferential rates is lost
Solution Approach 1:
User preferences, including negotiated preferential rates and quality of service criteria, are pre-configured and stored in the ranking machine before the service lookup occurs. When a user queries the directory, the ranking machine automatically applies these pre-set criteria to filter and rank services, ensuring that preferred services (such as those with negotiated rates) are prioritized without requiring the user to manually specify preferences during the lookup process.
3Productivity
If intermediate services are allowed to select lower-level services independently using a directory, then service invocation flexibility is improved, but control over service selection is lost to the originating requestor
Solution Approach 1:
The system implements a feedback mechanism where the ranking machine receives user preference information and continuously refines service selections based on this feedback. The ranking machine maintains a record of user preferences and uses this feedback to automatically adjust service rankings, ensuring that the originating requestor's preferences are consistently applied throughout the service invocation chain without requiring manual intervention at each level.
Data Source
AI summary
A method includes ranking services in a web services architecture and a web services architecture. The web services architecture has a hierarchy of services with a root originating service requester and a service of a first level in the hierarchy calling a service of a lower level. The originating service requestor indicates a preference regarding one or more services and a ranking machine provides a choice algorithm based on the preference. The originating service requestor invokes services of one or more levels of hierarchy. At each level of the hierarchy, a service uses a directory to find a set of possible lower-level services and the ranking machine applies the choice algorithm to the set of possible lower-level services. In one embodiment, the set of possible lower-level services is referred to the ranking machine from the directory and a preferred sequence is returned by the ranking machine to the directory.