Abstract
Cross-organizational worlflow appears to be more and more important in global business. It is a better solution to the problem of cross-organizational collaboration that state machine workflow is able to model business processes based on discrete points. Data will be transferred among different organizations in their collaborations and more human interactivities are needed. Further more, the event driven approach makes it more flexible to state transitions. Therefore, an approach for state machine workflow modeling based on event driven is presented. Using data-dependent methods, it can analyze and optimize the data exceptions to improve process efficiency and data accuracy before compiling workflow processes.