| Software testing nowadays plays a more and more important role in the software development cycle, for GUI software testing, including test path and test data, the test path and data selection is reasonable and comprehensive or not is a very important part of software testing, nowadays the automated test technology has been introduced to the GUI software testing filed,which to perform some work which need less manual participation,and repeated execution, but automatic script generation and maintenance still need a lot of extra artificial participation.The knowledge base is introduced in this paper. Firstly, the paper introduces the present situation and background of software test case generation method, analyzed their lack of treat measurement system in the form of expression, and is not conducive to the maintenance of test cases. In this paper, by introducing ontology model and inference rules to solve the above problem.Through the establishment of common software ontology model, including the constraint relationship between them, and the basic information of software and instantiate the measured software(vehicle management system) model, and use based on OWL ontology editor protege of the formalization of ontology expression. Then set up the rules of business logic, using SWRL language to express.Finally the paper convert the SWRL rules and OWL ontology to Jess rules and Jess facts and then reasoning test cases, then use the script generator based on Jena to automatic generate the scripting language, this paper selects the vehicle management system’s login process and gas flow of the automatic generation of test cases and test scripts, what’s more we analyze the quality of the results, and prove it in the rational robot platform. |