Research of knowledge expression model for Non-monotonic reasoning based on uncertain inference

Aiming to the application of Non-monotonic reasoning based on uncertain inference to plan recognition, a new way of knowledge expression model is proposed in mean of extending credibility of TMS node and building evidence transfer mechanism, which ensures the system to inference forward and build hypothesis under the case of the node state is noncommittal. The gene of node reliability and influence degree is introduced to overcome the shortage of only one value of credibility at one time and implement uncertain inference. Theoretical analysis and experimental results demonstrate that our method could describe clearly the process of non-monotonic reasoning based on uncertain inference, so it is expectable for our method to play a useful role in plan recognition system.