A Data Flow Testing Approach for Semantic Web Service Compositions Based on OWL-S

• Main Authors: Teng-Yi Huang, 黃騰毅
• Other Authors: Chien-Hung Liu
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/eg23e8

碩士 === 國立臺北科技大學 === 資訊工程系研究所 === 98 === Semantic Web Services (SWS) that integrate the concept of ontology to support incorporation of machine understandable semantics into web services has drawn much attention in recent years. Among the specification languages of SWS, OWL-S (Web Ontology Language for Service) is the most widely accepted standard proposed by W3C. OWL-S can be used to describe SWS and its associated semantic information, to define the operations of the SWS, and to select and compose loosely coupled SWS into a high-level composite service. However, OWL-S builds on the RDF (Resource Description Framework), OWL (Web Ontology Language), and SWRL (Semantic Web Rule Language) and is extremely complex and hard to understand. This makes OWL-S web services difficult to analyze and test. In view of this, this thesis proposes a data flow testing approach for web service compositions based on OWL-S. The test artifacts introduced by OWL-S are thoroughly analyzed. A test model is proposed to abstract the test information of OWL-S web services. The test model can facilitate testers to understand and analyze the control flow structures as well as the data flow dependencies and interactions of OWL-S web services. Based on the test model, algorithms are presented to compute data definition-use pairs and to derive test paths so as ensure the correctness of OWL-S web services. In addition, a semi-automatic tool is also developed to support the construction of the test model and the testing of OWL-S web services.