Contextual service discovery using term expansion and binding coverage analysis

• We propose a contextual approach to discovering web services.
• The approach mitigates issues of wording bias and ambiguity by term expansion.
• The approach ensures that discovered services are compatible with binding context.
• The approach can find out qualified services for constructing a composite service.

Business cloud emerges as a new solution for providing on-demand business services over Internet. Standardized service technologies drive widespread adoption of clouds and enable computation evolution towards service-oriented paradigm. While diversified cloud services are readily available today, how to discover desired services fitting into user’s context becomes a practical challenge. In this paper, we propose a Contextual Service Discovery (CSD) approach to help find out qualified services in accordance with binding context on the user side. Query descriptions and binding information are analyzed as a set of meaningful terms. We designed a term expansion mechanism to improve matchmaking performance by mitigating issues of wording bias and ambiguity. Besides, binding coverage analysis between multiple services and a given query is conducted to ensure that matched services are compatible with user’s contextual expectation. The experimental results show that the proposed approach is with better performance than other alternatives under Top-NN precision and recall metrics.