Evaluation of the Testing Methods in Agent-Oriented Software Engineering

Testing is an important process that can assure the quality and the correct functionality of the multi agent systems (MAS). Multiple testing methods in agent-oriented software engineering (AOSE) have been introduced in recent years. Although the quality of the system is dependent on the quality of the applied test method, very little attention has been paid to evaluating these testing methods. As a result, it is difficult to select a sufficient method for testing an agent-based system. Additionally, there are no means to determine what the advantages and drawbacks of each method are. This paper proposes a framework for evaluating and comparing the testing methods in AOSE. This framework addresses major divisions of a testing method. The framework is then used to evaluate some prominent testing methods, have been proposed so far. A subset of these testing methods, which cover more criteria in the proposed framework, is presented.