Semantic Web Services promise automatic service discovery and composition, relying heavily on domain ontology as a core component. With large Web Service repository, manual ontology development is proving a bottleneck (with associated expense and likely errors) to the realisation of a semantic Web of services. Providing the appropriate tools that assist in and automate ontology development is essential for a dynamic service vision to be realised. As a statement of research-in-progress, this paper proposes combining different ontology learning paradigms in Web Services domain, highlighting the need for further research that accommodates the variation in Web Service descriptive and operational sources. A research agenda is proposed that recognises this variation in artefacts as they are selected, pre-processed and analyzed by ontology learning techniques.