Prolog/Rex inference engine
Sanja Vraneš, Mladen Stanojević
Abstract
Sanja Vraneš, Mladen Stanojević
Abstract
Abstract: Prolog/Rex supports the construction of knowledge‐based systems, allowing frames (Prolog/Rex concepts) and rules to be combined, and providing language level support for hypothetical reasoning and assumption‐based truth maintenance (Prolog/Rex viewpoint mechanism). To increase run‐time performance during rule execution, Prolog/Rex provides two complementary rule compilation techniques: the simple indexing scheme and the heavily modified Rete expansion method. To improve the efficiency of the inference engine, we built an agenda‐based rule manager that uses priority mechanism and/or heuristics embedded in meta‐rules to control the rule firing order. Forward‐ and backward‐chaining can be combined within the same program.
OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Abstract: Prolog/Rex supports the construction of knowledge‐based systems, allowing frames (Prolog/Rex concepts) and rules to be combined, and providing language level support for hypothetical reasoning and assumption‐based truth maintenance (Prolog/Rex viewpoint mechanism). To increase run‐time performance during rule execution, Prolog/Rex provides two complementary rule compilation techniques: the simple indexing scheme and the heavily modified Rete expansion method. To improve the efficiency of the inference engine, we built an agenda‐based rule manager that uses priority mechanism and/or heuristics embedded in meta‐rules to control the rule firing order. Forward‐ and backward‐chaining can be combined within the same program.
Key concepts: Prolog, Computer science, Forward chaining, Backward chaining, Inference engine, Heuristics, Programming language, Search engine indexing