In open network environment, the loss of behavior footprint tokens producing in the interactions among the distributed software may lead unable to analyze and predict the software behavior. In order to find the event source for the lost token, translates the incomplete token problem of behavior footprint into the problem of network maxflow. Takes the along path forward of augmenting path algorithm to find the most possible behavior sequence for every transaction. The simulation experiment proves that the proposed method has validity and accuracy in marking the transaction of incomplete token.