Ontology is the key ingredient of semantic Web service technologies, which support systematic management of Web services, such as automatic service discovery, service composition, and change management. It is crucial and challenging to reduce the human efforts for developing ontologies. We propose a bottom-up approach that bootstraps operation-level service ontologies from WSDL descriptions. The approach leverages the techniques of information retrieval and machine learning. The relevance and similarity between Web services are measured based on the WSDL descriptions. The process of developing service ontologies consists of two steps. First, we build service ontologies based on the service relevance. We then construct the structure of the service ontologies based on the service similarity. We conduct an empirical study on real Web services to demonstrate the effectiveness of the proposed approach.