Beta-Prolog: An extended Prolog with Boolean tables for combinatorial searching

Most combinatorial problems, e.g. planning and constraint satisfaction problems, can be formulated as situation transition problems (STPs). Although Prolog is well used for searching problems, it is unsatisfactory for solving STPs due to its lack of appropriate data structures for representing situations. The authors describe an extended Prolog, called Beta-Prolog, that supports the definition and manipulation of Boolean tables. A Boolean table is a natural and efficient data structure for representing situations in STPs. With the primitives on Boolean tables, tests of conditions on situations and changes of situations can be performed in constant time. After defining Boolean tables, the authors describe programs and computational results for the following problems: transitive closure, blocks world planning, graph coloring and channel routing problems. They also describe briefly the implementation techniques of Beta-Prolog.