An abstract interpretation scheme for logic programs based on type expression

This paper describes a scheme for the abstract interpretation of logic programs based on type information. Our scheme has four components: generalisation, abstract unification, and summarisation are discussed. Our scheme does not suffer from the aliasing problem. Our main application of abstract interpretation is deriving producer-consumer relationships and classification of procedures for transformation of logic programs for efficient execution under committed-choice AND-parallelism extended to find all solutions. However, our scheme is general enough to be directly applied for type generation, compile time memory allocation and efficient unification