An optimized incremental attribute algorithm for the abstract interpretation of logic programs

We present a novel approach to implement an efficient, abstract interpretation algorithm for logic programs by means of attributed grammar. Successful implementation of this algorithm yielded memory storage saving and run-time reduced. We first execute the algorithm presented in [2] (slightly modified), which generates a. tree with four types of nodes. We then store in computer memory only the pertinent. subtrees. In other words, after each step, we prune some branchs of tree and develop only the subtree which can improve the "oracle" (i.e. set of abstract tuple).