The Plan Recognition Based On Probabilistic Goal Graph

Intelligent plan recognition is an hot branch of AI in resent years. As it' s widelyapplication, many famous scholars home and abroad devote themselves to the research.Especially, the incomplete and undetermined plan recognition problem is the key branchof this field. In various kinds of research approaches, because the probabilitymethod can describe the uncertain information ration relatively accurately, sostudy the probabilistic plan recognition that the operator has probabilisticoutcome relatively strong superiority in method. Many researchers are in favor ofthis method and have produced a large amount of algorithms on the basis of this.Goal Recognition Through Goal Graph Analysis is a excellent algorithm of these. In thispaper, we give extend the goal graph from the domain representation, goal graphconstruction etc. And then we give the new definition of causal link intensity, relevantaction supply intensity, consistent goal intensity and so on, using which we give the newalgorithm of probabilistic plan recognition based on probabilistic goal graph. Ouralgorithm not only keeps down the virtue of goal graph, but also extends the goal graphto handle the probabilistic plan recognition problems. As our algorithm arises, thealgorithm of probabilistic plan recognition based on probabilistic goal graph canapplication in many fields.