An Essential Hybrid Reasoning System: Knowledge and Symbol Level Accounts of KRYPTON

Hybrid inference systems are an important way to address the fact that intelligent systems have muiltifaceted rep-resentational and reasoning competence. KRYPTON is an experimental prototype that competently handles both terminological and assertional knowledge; these two kinds of information are tightly linked by having sentences in an assertional component be formed using structured complex predicates denned in a complementary terminological component. KRYPTON is unique in that it combines in a completely integrated fashion a frame-based description language and a first-order resolution theorem-prover. We give here both a formal Knowledge Level view of the user interface to KRYPTON and the technical Symbol Level details of the integration of the two disparate components, thus providing an essential picture of the abstract function that KRYPTON computes and the implementation technology needed to make it work. We also illustrate the kind of complex question the system can answer. I Introduction Many of today's knowledge representation (KR) systems offer their users a choice of more than one language for expression of domain knowledge. While the idea has been important to the field for many years (e.), "multiple representations" seems to have recently become a popular catch phrase. Many of the modern expert system development environments wave the polyglot banner , and except perhaps for some stalwart first-order logicians, most everyone would probably agree that one uniform language will not serve all representational needs. It is sometimes difficult to discern the true value of multiple languages; some of the commercial development tools seem simply to appeal to "the more the merrier," without any clear idea of how merrier is better. However, on the research front, there have fortunately been some coherent views expressed on the merits of bringing disparate dialects together. The arguments have mainly to do with the naturalness of expressing certain kinds of facts in certain forms (see, e.g., [Rich, 1982]), or with the efficiency of computing some inferences once some datum is massaged into a certain representational form [Genesereth, 1981]. 2 The arguments in favor of naturalness and computational superiority are important ones. However, as we have argued elsewhere [Brachman and Levesque, 1982], there may be a simpler, more basic reason to diversify: an intelligent system has more than one kind of representational need. For example, it is interesting to observe that so far, most KR systems only allow the expression of what we would call assertional knowledge— statements of facts or beliefs. If …