Verification of problem-based learning systems using modified petri nets

In this paper, we intend to verify a web-based system on problem-based learning (PBL). To consider the design flow of the web-based PBL system, it is essential to avoid the potential hazard introduced by a logically incorrect system design. In order to eliminate the potential hazard, we define a new class of Petri net, namely, an Activity Flow (AF) net, which is suitable to be converted from a UML (Unified Modeling Language) activity diagram. Through the siphon-based deadlock detection of the AF net, we can find whether there is a process hazard in the UML activity diagram or not. This is helpful to implement the PBL system and to ensure the correct activities and the right control flow. In addition, we attempt to enhance the quality of the system verification by using a questionnaire. Thus we can interpret the user's level of satisfaction with the designed PBL system. These two verification approaches bring us to achieve an adequately positive response to the web-based PBL system. (C) 2012 Elsevier Ltd. All rights reserved.