Matthew only focused on deep learning as artificial intelligence. Researchers in ai have produced many theories of non monotonic reasoning that be seen also as attempting to give a semantics for genericity. What is monotonic and nonmonotonic reasoning in artificial. Various systems of nonmonotonic reasoning have been developed, among which are nonmonotonic logic, circumscription.
Prehistory of relations nonmonotonic reasoning is considered today an essential part of the logical approach to arti cial intelligence. Nonmonotonic reasoning in description logic by tableaux. Monotonic and nonmonotonic reasoning in artificial. Artificial intelligence 33 non monotonic reasoning in ai. The term nonmonotonic logic in short, nml covers a family of formal.
Nonmonotonic logic stanford encyclopedia of philosophy. It has always been of very much interest of men to explain cognition and reasoning. As we shall see, a defaulttheory may have zero, one. For example when we visit a friends home, we buy biscuits for the children. An overview of nonmonotonic reasoning and logic programming.
In common sense reasoning one often draws conclusions that have to be withdrawn when further information is obtained. Logic will be said as non monotonic if some conclusions can be invalidated by adding more knowledge into our knowledge base. In particular, it provides rigorous mechanisms for taking back conclusions that, in the presence of new information, turn out to be wrong and for deriving new, alternative conclusions instead. Put another way, ai researchers aim to construct a synthetic homo economicus, the mythical perfectly rational agent of neoclassical economics. Moore artificial intelligence center sri international, menlo park, ca 94025 abstract commonsense reasoning is nonmonotonic in the sense that we often draw conclusions on the basis of partial information that we retract when we are given more complete information. If artificial intelligence ai researchers can agree on anything, it is that. This is the equivalent of diagnostic reasoning in bayesian networks ocw, 2005. In this article we introduce a non monotonic reasoning engine, i.
It is based, in part, on a discussion of intelligent agent architecture in 12, chapter. Laboratoire cnrs lsis 20102011a ai am indebted to laura giordano and alberto martelli for having provided me their course material. The definite clause logic is monotonic in the sense that anything that could be concluded before a clause is added can still be concluded after it is added. In other words, non monotonic logics are devised to capture and represent defeasible inferences cf. The definite clause logic is monotonic in the sense that anything that could be concluded before a clause is added can still be. Nonmonotonic reasoning is useful for representing defaults. In this paper we introduce nonmonotonic reasoning and the attempts at formalizing it using nonmonotonic logics.
Schaub bpl a school of computing science, simon fraser university, burnaby, bc, canada vsa ls6. Hence a nonmonotonic inference may involve the theory as a whole. Tu dresden, ws 201718 introduction to nonmonotonic reasoning slide 51. Nonmonotonic reasoning, nonmonotonic logics and reasoning. Nonmonotonic reasoning, conditional objects and possibility. Nonmonotonic reasoning in description logic by tableaux algorithm with blocking jarom r malenko and petr st ep anek charles university, malostranske namesti 25, 11800 prague, czech republic, jaromir. Nerode classical logic is the study of safe formal reasoning. The objective of the article is to present solutions for problems occurring in classical expert systems based on.
Classical logics share a common property of being monotonic. Each of the several papers that introduces a mode of nonmonotonic reasoning seems to have a particular application in mind. This is called default or nonmonotonic reasoning and cannot be. Before proceeding to applications of circumscription i want to suggest a typology of the uses of nonmonotonic reasoning. Monotonic reasoning can be defined in artificial intelligence. Perhaps we are looking at different parts of an elephant. Makinson and gardenfors 1991 david makinson, pe ter gardenfors, relations between the logic of theory change and nonmonotonic logic, in.
Adding knowledge in monotonic reasoning does not reduce the set of prepositions that arise. Ai or artificial intelligence is the simulation of human intelligence processes by machines, especially computer systems. I will give a simplified account of this view here. Default or nonmonotonic reasoning in the previous example we made an assumption in the absence of information to the contrary, and revised this assumption later when new knowledge became available. Nonmonotonic reasoning computer science university of.
Nonmonotonic reasoning essays celebrating its 30th. Non monotonic reasoning attempts to formalize reasoning with ab kr incomplete information by classical logic systems. Nonmonotonic reasoning provides formal methods that enable intelligent systems to operate adequately when faced with incomplete or changing information. Monotonic reasoning artificial intelligence definition. Horty philosophy department and institute for advanced computer studies, university of maryland, college park, md 20742, usa richmond h. S is assumed to be true as long as there is no evidence to the contrary. Coskun sonmez reasoning iintroduction as studies of artificial intelligence continue, it should become apparent that progres in solving the problems of ai closely parelleled the development of tools and technics for manipulating knowladge. Commonsense reasoning 37 write down axiomatize what everyone knows about some domain for example, space shuttles, and apply a formal proof procedure for this axiomatization as the machinery of rea soning.
The field of artificial intelligence ai strives to build rational agents capable of perceiving the world around them and taking actions to advance specified goals. In non monotonic reasoning, some conclusions may be invalidated if we add some more information to our knowledge base. The objective behind the area is the use of logic for knowledge. Nonmonotonic reasoning in artificial intelligence semantic scholar. Introduction to non monotonic reasoning master recherche sis, marseille. Semantical considerations on nonmonotonic logic robert c. A rule like ps are normally, usually qs is commonly referred to as a default, and the goal is to account for default reasoning not to be confused with default logic, which is a speci c approach. The system associates facts asserted with the rules and facts used to justify them and as those facts change updates the dependent knowledge as well. A directly cautious theory of defeasible consequence for default logic via the notion of general extension. This volume contains the papers that will be presented at the first international workshop on argumentation and nonmonotonic reasoning argnmr 2007 on may 14, 2007 in tempe, arizona, us. The standard example in ai of a nonmonotonic reasoning pattern has to do with flying. I discuss the prehistory of nonmonotonic reasoning starting from approximately 1958.
We offer such a semantics here for one kind of genericity. For this article we discuss the area of logicbased ai and in particular non monotonic reasoning. Artificial intelligence foundations of computational. Non monotonic reasoning allows various kinds of hypothetical reasoning. Mueller says that commonsense reasoning is the sort of reasoning we all perform about the everyday world. Elsevier artificial intelligence 77 1995 3257 artificial intelligence on the acceptability of arguments and its fundamental role in nonmonotonic reasoning, logic programming and nperson games phan minh dung division of computer science, asian institute of. A logic is non monotonic if some conclusions can be invalidated by adding more knowledge. Western philosophers developed classical logic over a period of thirtythree centuries after its introduction in the form of syllogistic by aristotle 1 in the third century b. In nonmonotonic theories, an inference may depend on lack of information. This barcode number lets you verify that youre getting exactly the right version or edition of a. The area of nonmonotonic reasoning nmr in arti cial intelligence ai studies such reasoning. Several paraphrases of this pattern are commonly accepted.
The human brain can solve highly abstract reasoning problems using a neural network that is entirely physical. Nmr then is a central and crucial area of ai, and is fundamental to commonsense reasoning. Nonmonotonic reasoning nonmonotonic reasoning deals with the problem of deriving plausible conclusions, but not infallible, from a knowledge base a set of formulas. Elsevier artificial intelligence 90 1997 3048 artificial intelligence compiling specificity into approaches to nonmonotonic reasoning james p. Nonmonotonic reasoning attempts to formalize reasoning with ab. Commonsense reasoning is very complex and it requires a large amount of knowledge about the world and the ability to use that knowledge. Artificial intelligence computer science and information.
The focus of this paper is nonmonotonic reasoning as it relates to logic programming. The logic of definite clauses with negation as failure is non monotonic. Logic and artificial intelligence 35 assumptions about what we take the real world to be and about how our machines interact with that world. In this chapter we show that it is possible to set up a logical model of databases using subimplication and two sets of stable formulas. Compiling specificity into approaches to nonmonotonic. Elsevier artificial intelligence 77 1995 3257 artificial intelligence on the acceptability of arguments and its fundamental role in nonmonotonic reasoning, logic programming and nperson games phan minh dung division of computer science, asian institute of technology, gpo box 2754, bangkok 10501, thailand. Tu dresden, ws 201718 introduction to nonmonotonic reasoning slide 47. A system of nonmonotonic reasoning, called subimplication, has been defined in bossu and siegel 1981. This is called default or non monotonic reasoning and cannot be formalized directly in pure fol in general. Oct 12, 2016 the human brain can solve highly abstract reasoning problems using a neural network that is entirely physical. In 1980, the artificial intelligence journal published an issue vol. The past 30 years have seen much impressive and important work in nmr, beginning with the seminal arti cial intelligence journal issue on the topic aij, 1980. In other words, nonmonotonic logics are devised to capture and represent defeasible inferences cf. Statistical reasoning a foundation for semantic web reasoning.
Nonmonotonic reasoning is a particular kind of plausible reasoning. Abductive reasoning allows a system to infer the possible causes for a certain effect. A skeptical theory of inheritance in nonmonotonic semantic. These processes include learning the acquisition of information and rules for using the information, reasoning using the rules to reach approximate or definite conclusions, and selfcorrection. The set of conclusions thus does not grow monotonically with the given information. Artificial intelligence 311 a skeptical theory of inheritance in nonmonotonic semantic networks john f. Readings in nonmonotonic reasoning illustrated edition by matthew ginsberg author isbn. A nonmonotonic logic is a formal logic whose consequence relation is not monotonic. Compiling specificity into approaches to nonmonotonic reasoning. Volume, issues 12, pages 1174 april 1980 download full issue.
Default or non monotonic reasoning in the previous example we made an assumption in the absence of information to the contrary, and revised this assumption later when new knowledge became available. For example, the possible courses for learning artificial intelligence at mit are 6. The underlying mechanisms are only partially understood, but an artificial network. Anything can end before a clause is added, even after it is added. Nonmonotonic reasoning is a subfield of artificial intelligence trying to find more realistic formal models of reasoning than classical logic. Full text of the second edition of artificial intelligence.
Nonmonotonic reasoning and multiple belief revision. Despite the difference in motivation and technical construction there are strong similarities between these logics which are confirmed when they are finally shown to have a common basis. A non monotonic logic is a formal logic whose consequence relation is not monotonic. Non monotonic reasoning deals with incomplete and uncertain models. Virtually every example in ai that calls upon such reasoning fits the following pattern. Non monotonic reasoning is based on default reasoning or most probabilistic choice. Thomason intelligent systems program, university of pittsburgh.