Building Web Service Ontologies
R.M. Sabou · Data Archiving and Networked Services (DANS) · 2006
The central question investigated by this thesis is:How to facilitate the process of building Web service ontologies and how to enhance the quality of these ontologies?Our work was centered around the following concrete research questions: Q1. What requirements should Web service ontologies fulfill?A prerequisite for improving the state of the art in Web service ontology development is to have a good understanding of this process.In particular, it is important to identify the requirements that Web service ontologies should fulfill. Q2. How to enhance the quality of generic Web service ontologies?Since generic Web service ontologies are used as standards for creating semantic descriptions they should have a high quality.To ensure this we need ways to test if they fulfill the previously identified requirements, to identify problematic aspects or limitations and to provide appropriate solutions. Q3. Is semi-automatic acquisition of Web service domain ontologies feasible?The acquisition of Web service domain ontologies is a time consuming task whose automation is desirable.It is important to identify the requirements for an automatic acquisition solution in the context of Web services and the data sets that are typically available.Existing ontology learning methods have to be adapted to the particularities of the context, rigorously evaluated and incorporated in easy to use tools.Our work on these research questions lead to several contributions, as summarized in the next section.