Showing papers on "Abductive reasoning published in 1996"

Connectionist-Symbolic Integration: From Unified to Hybrid Approaches

Abstract: Contents: Preface. R. Sun, An Introduction to Hybrid Connectionist-Symbolic Models. Part I: Reviews and Overviews M. Hilario, An Overview of Strategies for Neurosymbolic Integration. R. Khosla, T. Dillon, Task Structure Level Symbolic-Connectionist Architecture. Y. Lallement, F. Alexandre, Cognitive Aspects of Neurosymbolic Integration. L. Magdalena, A First Approach to a Taxonomy of Fuzzy-Neural Systems. Part II: Learning in Multi-Module Systems T. Johnson, J. Zhang, H. Wang, A Hybrid Learning Model of Abductive Reasoning. R. Sun, T. Peterson, A Hybrid Agent Architecture for Reactive Sequential Decision Making. M. Malek, B. Amy, A Preprocessing Model for Integrating Case-Based Reasoning and Prototype-Based Neural Network. B. Orsier, A. Labbi, A Step Toward Fully Integrated Hybrid Systems. J.C. Gonzalez, J.R. Velasco, C.A. Iglesias, A Distributed Platform for Symbolic-Connectionist Interoperation. Part III: Representing Symbolic Knowledge. B. Kokinov, Micro-Level Hybridization in the Cognitive Architecture Dual. S. Stevenson, An Integrated Symbolic/Connectionist Parsing Architecture. X. Wu, M. McTear, P. Ojha, H. Dai, A Hybrid System Framework for Disambiguating Word Senses. N.S. Park, D. Robertson, A Localist Network Architecture for Logical Inference. J. Austin, A Distributed Associative Memory for Symbolic Reasoning. E. Mjolsness, Symbolic Neural Networks Derived From Stochastic Grammar Domain Models. Part IV: Acquiring Distributed Representation. T.A. Plate, Structure Matching and Transformation With Distributed Representations. A. Sperduti, A. Starita, C. Goller, Distributed Representations for Terms in Hybrid Reasoning Systems. R. Krosley, M. Misra, Symbolic Distributed Representations. Part V: Epilog F. Alexandre, An Analysis of Connectionist-Symbolic Integration.

On Van Fraassen's Critique of Abductive Reasoning

Abstract: Adoptant l'attitude critique propre a A. Menuge, l'A. montre que ni l'argument du mauvais lot, ni l'argument de l'indifference, ni la conception de la non-observabilite, developpes par Van Fraassen contre le raisonnement adbuctif, ne parviennent a refuter le principe du privilege et l'approximation de la verite qui fondent l'empirisme constructif

Abduction: A Theory of Visual Interpretation

Abstract: Peirce's notion of abductive reasoning provides a theoretical framework in which to analyze visual interpretation, that is, how viewers understand a visual and interpret its meaning. This paper demonstrates that a different kind of interpretive logic operates for visual communication processes than for language-based communication processes, and this logic is best articulated in the semiotic literature where the notion of interpretation is more carefully conceptualized.

Visual Abductive Reasoning in Archaeology

Abstract: Biographical studies have shown that visual mental imagery plays a significant role in the conduct of scientific research, particularly in the generation of hypotheses. But the nature of visual mental imagery and its participation in abductive inference is not systematically understood. This paper discusses examples of visual abductive reasoning by archaeologists, analyzing them according to the visual information and the process of inference employed. This work supports the conclusion that visual abduction is useful to scientists under certain conditions and that it is amenable to detailed study.

The Role of Abductive Reasoning within the Process of Belief Revision

Abstract: An inquiring agent is concerned with obtaining as much new, error-free, information as possible. One way of doing this is to simply incorporate information presented to an agent as is. This strategy is adopted by many belief revision frameworks including the popular AGM framework. A more natural strategy would be for the agent to first seek an explanation or justification for the new information. After doing so, it could incorporate the explanation into its epistemic state together with the new information. Such a strategy would be particularly effective if the agent’s situation does not allow it to obtain new information easily. We model this strategy through the use of abductive reasoning. This allows us to then investigate the role of abductive inference within a belief revision framework based on the AGM. We not only look at the incorporation of new information but also at the removal of information. We begin by looking at some logical aspects of abduction and to contrast it, in a pragmatic sense, with the process of induction as performed by inverse resolution. We proceed to develop an account of an abductive expansion operator in the vein of the AGM framework. A definition, postulates and several constructions, reminiscent of the AGM, are developed together with a number of representation theorems. It is also shown how abductive expansion is related to nonmonotonic inference, in particular, default reasoning. The process of contraction is then investigated and we note how abduction can already be viewed as an active part of this operation. However, abductive expansion and AGM contraction do not exhibit the dual behaviour one might expect. This leads us into an investigation of an alternate operation known as Levi-contraction. We suggest a Grove style semantic modelling and provide additional postulates in order to obtain a complete characterisation. Our emphasis on expansion and contraction is guided to a large extent by Levi’s commensurability thesis which states that any revision can be achieved through a series of expansion and contraction operations. However, using our work on expansion and contraction, we briefly investigate the repercussions for an abductive revision operator determined through the Levi identity. It turns out that this problem relies heavily on that of iterated revision.

Using compiled knowledge to guide and focus abductive diagnosis

Abstract: Several artificial intelligence architectures and systems based on "deep" models of a domain have been proposed, in particular for the diagnostic task. These systems have several advantages over traditional knowledge based systems, but they have a main limitation in their computational complexity. One of the ways to face this problem is to rely on a knowledge compilation phase, which produces knowledge that can be used more effectively with respect to the original one. We show how a specific knowledge compilation approach can focus reasoning in abductive diagnosis, and, in particular, can improve the performances of AID, an abductive diagnosis system. The approach aims at focusing the overall diagnostic cycle in two interdependent ways: avoiding the generation of candidate solutions to be discarded a posteriori and integrating the generation of candidate solutions with discrimination among different candidates. Knowledge compilation is used off-line to produce operational (i.e., easily evaluated) conditions that embed the abductive reasoning strategy and are used in addition to the original model, with the goal of ruling out parts of the search space or focusing on parts of it. The conditions are useful to solve most cases using less time for computing the same solutions, yet preserving all the power of the model-based system for dealing with multiple faults and explaining the solutions. Experimental results showing the advantages of the approach are presented.

Prolegomena to Logic Programming for Non-monotonic Reasoning

Abstract: The present prolegomena consist, as all indeed do, in a critical discussion serving to introduce and interpret the extended works that follow in this book As a result, the book is not a mere collection of excellent papers in their own specialty, but provides also the basics of the motivation, background history, important themes, bridges to other areas, and a common technical platform of the principal formalisms and approaches, augmented with examples

An Abductive Framework for Negation in Disjunctive Logic Programming

Abstract: In this paper, we study an abductive framework for disjunctive logic programming that provides a new way to understand negation in disjunctive logic programming. We show that the defined framework captures the existing minimal model semantics based on (Extended) Generalised Closed World Assumption ((E)GGWA), This relationship between abduction and minimal model reasoning provides a methodology to develop algorithms for minimal model reasoning. To demonstrate this, we show how a theorem prover, based on restart model elimination calculus, can be modified for abductive reasoning and thus for minimal model reasoning.

System and methods for intelligent database management using abductive reasoning

Abstract: A system of the present invention includes a computer having a processor and a memory, which includes general-purpose components providing support for application tasks processing. In a preferred embodiment, the general-purpose components include a Database Management System (50), a Deductive Database (40), and an Abductive Metainterpreter (30); the latter including an engine for abducing new information from known facts stored in the database. In this fashion, a user may model an environment or domain of interest (e.g., a communication network). Once defined, the environment is automatically maintained by the system of the present invention. In response to a user request for modifying the environment, for example, the system automatically generates transactions which satisfy the request.

Meta-Cases: Explaining Case-Based Reasoning

Abstract: AI research on case-based reasoning has led to the development of many laboratory case-based systems. As we move towards introducing these systems into work environments, explaining the processes of case-based reasoning is becoming an increasingly important issue. In this paper we describe the notion of a meta-case for illustrating, explaining and justifying case-based reasoning. A meta-case contains a trace of the processing in a problem-solving episode, and provides an explanation of the problem-solving decisions and a (partial) justification for the solution. The language for representing the problem-solving trace depends on the model of problem solving. We describe a task-method-knowledge (TMK) model of problem-solving and describe the representation of meta-cases in the TMK language. We illustrate this explanatory scheme with examples from Interactive Kritik, a computer-based design and learning environment presently under development.

Irrelevance and conditioning in first-order probabilistic logic

Abstract: First-order probabilistic logic is a powerful knowledge representation language. Unfortunately, deductive reasoning based on the standard semantics for this logic does not support certain desirable patterns of reasoning, such as indifference to irrelevant information or substitution of constants into universal rules. We show that both these patterns rely on a first-order version of probabilistic independence, and provide semantic conditions to capture them. The resulting insight enables us to understand the effect of conditioning on independence, and allows us to describe a procedure for determining when independencies are preserved under conditioning. We apply this procedure in the context of a sound and powerful inference algorithm for reasoning from statistical knowledge bases.

An Abductive Proof Procedure for Reasoning About Actions in Modal Logic Programming

Abstract: In this paper we propose a modal approach for reasoning about actions in a logic programming framework. We introduce a modal language which makes use of abductive assumptions to deal with persistency, and provides a solution to the ramification problem, by allowing one-way “causal rules” to be defined among fluents.

Approximate and Commonsense Reasoning: From Theory to Practice

Abstract: This paper provides an overview of present trends in approximate and commonsense reasoning. The different types of reasoning, which can be covered by this generic expression, take place when the available information is either incomplete, or inconsistent, or pervaded with uncertainty, or imprecise and qualitative. The conclusions which are then obtained are usually plausible but uncertain. Yet, approximate or commonsense reasoning is useful in practical problems such as prospect evaluation, diagnosis, forecasting and decision tasks, where better information cannot be got. Classical logic is insufficient for handling these types of reasoning. Different ideas of orderings play a role in these reasoning processes: plausibility orderings between interpretations or situations which are unequally uncertain, similarity orderings with respect to prototypical situations or cases, preference orderings between acts or situations when the problem is a matter of choice. These orderings can be encoded using purely ordinal scales, or scales with a richer structure (when it is meaningful and compatible with the quality of the available information). This general idea of ordering provides a kind of unification between the different reasoning modes and somewhat typifies approximate and commonsense reasoning. Advances in default reasoning, inconsistency handling, data fusion, updating, abductive reasoning, interpolative reasoning, and decision issues in relation with Artificial Intelligence research, are briefly reviewed. Open questions and directions for future research which seem especially important for the development of practical applications are pointed out. The paper is largely based on authors' research experience, and as such, presents a rather personal view, which may not be exempt from some biases.

A Case Study in Reasoning about Actions and Continuous Change.

Abstract: This paper shows how the Situation Calculus can be extended to deal both with `narratives' and with domains containing real-valued parameters, whose actual values may vary continuously between the occurrences of actions. In particular, a domain is represented where action occurrences may be `triggered' at instants in time when certain parameters reach particular values. Its formalisation requires the integration of several types of default reasoning. Hence Baker's circumscriptive solution to the frame problem is extended to re ect the assumptions that by default a given action does not occur at a given time point, that by default a given set of parameter values does not trigger a given action, and that by default a given action occurrence does not result in a discontinuity for a given parameter. Regarding the minimisation of discontinuities, the example illustrates how circumstances can arise where, at a particular time point, discontinuities in some parameters can be `traded' for discontinuities in others. It is argued that, in general, in such cases extra domain-speci c information will be necessary in order to eliminate anomalous models of

Abductive change operators

Abstract: This paper describes a change theory based on abductive reasoning. We take the AGM postulates for revisions, expansions and contractions, and Katsuno and Mendelzon postulates for updates and incorporate abduction into them. A key feature of the theory is that presents a unified view of standard change operators and abductive change operators rather than a new and independent change theory for abductive changes. Abductive operators reduce to standard change operators in the limiting cases.

An evolutionary computing approach to probabilistic reasoning on bayesian networks

Abstract: Bayesian belief networks can be used to represent and to reason about complex systems with uncertain or incomplete information Bayesian networks are graphs capable of encoding and quantifying probabilistic dependence and conditional independence among variables Diagnostic reasoning, also referred to as abductive inference, determining the most probable explanation (MPE), or finding the maximum a posteriori instantiation (MAP), involves determining the global most probable system description given the values of any subset of variables In some cases abductive inference can be performed with exact algorithms using distributed network computations, but the problem is NP-hard, and complexity increases significantly with the presence of undirected cycles, the number of discrete states per variable, and the number of variables in the network This paper describes an approximate method composed of a graph-based evolutionary algorithm that uses nonbinary alphabets, graphs instead of strings, and graph operators to perform abductive inference on multiply connected networks for which systematic search methods are not feasible The motivation, basis, and adequacy of the method are discussed, and experimental results are presented

Contextual reasoning is NP-complete

Abstract: The logic of context with the ist (c, p) modality has been proposed by McCarthy as a foundation for contextual reasoning. This paper shows that propositional logic of context is NP-complete and therefore more tractable than multimodal logics or Multi Language hierarchical logics which are PSPACE-complete. This result is given in a proof-theoretical way by providing a tableau calculus, which can be used as a decision procedure for automated reasoning. The computational gap between logic of context and modal logics is analyzed and some indications for the use of either formalisms are drawn on the basis of the tradeoff between compactness of representation and tractability of reasoning.

3 Hypotheses and Geomorphological Reasoning

Abstract: Geomorphology is a way of thinking about the surface of planet Earth. Controlled experimentation, in the manner of pure physics, is not possible for most geomorphological concerns. Thus, much of conventional analytical philosophy of science, which is based on the exemplar of experimental physics, fails to portray important aspects of geomorphological reasoning. This is particularly true of hypothesizing, which was recognized by Gilbert, Chamberlin, and Davis as a central methodological concern of geomorphology. Geomorphological reasoning largely relies upon retroductive inference, which Charles S. Peirce described as 'the spontaneous conjectures of instinctive reason'. Because it reasons from real effects to real causes, eventually colligating (bring together) facts under a conceptual scheme (hypothesis), retroduction bridges the gulf between nature and mind. Geomorphological indices, such as landforms and sediments, are signs for which causative processes are inferred retroductively. Though superficially similar to lucky , 'guessing', retroductive inference succeeds in generating fruitful hypotheses (some of them outrageous) because the human mind is instinctively attuned to certain aspects of nature. This instinctive propensity in science to 'guess right', which Galileo called il lume naturale, may derive from fundamental properties of the universe and mind that modern cosmologists have named the 'anthropic principle'.

Realism, history and the quantum theory: Philosophical and historical arguments for realism as a methodological thesis.

Abstract: Scientific realists and non-realists disagree over the reach of scientific knowledge: does it extend beyond the observational realm. Intuitions about abductive inferences are at the heart of many realist positions, but are brought into question by the non-realists' contention that theories are underdetermined by data, and the alleged circularity of realist attempts to show that such inferences are reliable. Some realists have tried to circumvent this problem by constructing methodological arguments for realism: if realism is embedded in scientific practice, the realist's picture of science might provide the best explanation of scientific success. Some non-realists reply by again pointing to the circularity of this strategy, which relies, again, on an abductive inference. Others deny that scientists do adopt realist stances. A methodological realist position is constructed: realist constraints on the acceptance and pursuit of theories-for instance requirements of intertheoretic coherence, and the avoidance of ad hoc explanation-have often contributed to progress in science. The position is immune to non-realist worries about the circularity of realist arguments, for it is a thesis about how science is practised, not the kind of knowledge it provides. The argument is pursued within a diachronic account of theory appraisal: Imre Lakatos' methodology of scientific research programmes (MSRP) examines the principles that govern the construction of theories, and provides criteria-achievement of progress-for the appraisal of research programmes. Although Lakatos may have seen these selection criteria, when fulfilled, as symptoms of something else-the fulfilment in the theory's development of some ideal of scientific honesty-achievement of Lakatosian progress can Serve as an end in itself. The realist methods mentioned in the last paragraph are then appraised as means to this end. Since the position has a methodological formulation and background, it is applied as a historical thesis to case studies in line with Lakatos' metamethodology. These comprise two explanatory forays into history: the consistency of Bohr's 1913 model of the atom, and the construction by Heisenberg and Schrodinger of the two original formulations of quantum mechanics. There follows one contemporary application: the construction of explanations in quantum chemistry using approximate models of molecules.

An information-base system based on the self-organization of concepts represented by terms

Abstract: Since multimedia information is complicated inform and vast in amount, conventional database-management systems or knowledge-base-management systems are hardly appropriate to store, manage, and utilize expertise effectively A new type of information model is developed according to an analysis of the information used by specialists for research and development, and a prototype information-management system is implemented The system consists of three parts: (1) flexible storage without special constraints on format and representation; (2) self-organization of terms by extracting semantic relationships among them; and (3) advanced utilization functions such as analogical reasoning, inductive inference, abductive inference, as well as information retrieval, numerical calculation, and deductive inference Thesauri which are automatically compiled and refined are used as conceptual structures of the information Thus obtained, conceptual structures can be used for sophisticated applications, including analogical reasoning, induction, and abduction The principle of open-world reasoning and an algorithm of analogy are developed An example of practical application to polymer information is presented

Parallel cost-based abductive reasoning for distributed memory systems

Abstract: This paper describes efficient parallel first-order cost-based abductive reasoning for distributed memory systems A search control technique of parallel best-first search is introduced into abductive reasoning mechanism, thereby finding much more efficiently a minimal-cost explanation of a given observation We propose a PARallel Cost-based Abductive Reasoning system, PARCAR, and give an informal analysis of PARCAR We also implement PARCAR on an MIMD distributed memory parallel computer, Fujitsu AP1000, and show some performance results

A connectionist framework for reasoning: reasoning with examples

Abstract: We present a connectionist architecture that supports almost instantaneous deductive and abductive reasoning. The deduction algorithm responds in few steps for single rule queries and in general, takes time that is linear with the number of rules in the query. The abduction algorithm produces an explanation in few steps and the best explanation in time linear with the size of the assumption set. The size of the network is polynomially related to the size of other representations of the domain, and may even be smaller. We base our connectionist model on Valiant's Neuroidal model (Val94) and thus make minimal assumptions about the computing elements, which are assumed to be classical threshold elements with states. Within this model we develop a reasoning framework that utilizes a model-based approach to reasoning (KKS93; KR94b). In particular, we suggest to interpret the connectionist architecture as encoding examples of the domain we reason about and show how to perform various reasoning tasks with this interpretation. We then show that the representations used can be acquired efficiently from interactions with the environment and discuss how this learning process influences the reasoning performance of the network.

Translating Case-Based Reasoning into Abductive Logic Programming.

Abstract: This paper presents a translation method of casebased reasoning which changes similarity according to context into abductive logic programming based on a generalized stable model semantics [5]. This kind of dynamic similarity can be found in CBR systems for legal reasoning such as HYPO [1]. Abductive logic programming is suitable to implement this dynamic similarity by regarding abducible predicates as similarity and changing abducible predicates by context. In this paper, we de ne a relevance criteria of cases for defendant and plainti and show how to change similarity between cases according to its context (defendant vs plainti , the current case and the cited case). We show correspondence between properties in a case used for similarity and abducibles used in a translated abductive logic program and show that we can construct an argument by using abducibles which explains why the current case is similar to the cited case and the current case is not similar to every counter case.

Abductive inference of events: diagnosing cardiac arrhythmias

Abstract: The ability to reason abductively, to infer a set of causes from their observed effects, is an essential component of human intelligence, and is used to construct explanations in such diverse areas as archaeology, criminal investigation, and medicine. Automating this process is, therefore, an important problem in artificial intelligence, but also an extremely challenging one. Its main difficulty lies in the exponential number of explanations for even the simplest set of observations. Abductive inference for real-world problems has been shown to be NP-hard. We offer a new approach to this problem--that of controlling the combinatoric explosion of possible explanations by means of temporal constraint propagation. To show the effectiveness of this approach, we have chosen to generate causal explanations of a cardiac arrhythmia appearing on an electrocardiogram (ECG) Explanations in this domain are constrained by well-understood temporal constraints from the ECG, as well as more subtle ones imposed by the underlying model of the heart. We demonstrate this approach with H scOLMES, an abductive reasoning system augmented by a method for propagating temporal constraints, which yields early refutation of inconsistent hypotheses. While H scOLMES is designed as a general-purpose system, it is used here to generate causes of cardiac arrhythmia. Though other arrhythmia diagnostic systems discover more arrhythmias than H scOLMES, they do so at a cost, in that they incorporate huge amounts of domain knowledge into their code, requiring a considerable investment of time and expertise. H scOLMES, on the other hand, provides very detailed explanations from a network of components embodying sparse amounts of domain knowledge and constrained by a weak set of temporal constraints. While H scOLMES generates more detailed explanations than are achieved in comparable diagnostic systems, its principal contribution lies in the alternative it offers to the traditional expert system/knowledge engineering approach to diagnostic problems. Many problems formerly thought to require immense amounts of domain-specific knowledge can, in fact, be solved efficiently by loosely connected networks of not very knowledgeable agents, provided that they are grounded in a good causal model of the problem.

A Temporal Extension to the Parsimonious Covering Theory

Abstract: In this work we propose a temporal extension to the Parsimonious Covering Theory (PCT). PCT provides a theoretical foundation for the diagnostic reasoning process as an abductive reasoning based associations between causes with their consequences. Our temporal extension of PCT allows one to associate to a disease a temporal evolution of its symptoms.

Abductive inferences and the structure of scientific knowledge

Abstract: The received theories of epistemology identify abductive inferences with the cognitive patterns of speculation (hypothesis formation) and insist that they cannot verify or confirm hypotheses. I criticize various descriptions of abduction, offer a structural analysis of abductive inference,, characterize abduction without alluding to its putative role in inquiry, and then demonstrate that some abductions do provide evidence and that not all scientific hypotheses derive from abductive inferences. This result challenges those notions of scientific k knowledge that dismiss some central scientific ideas (for example, evolution) as ‘meta-physical research programs’ or ‘just theories’ when they are instead well substantiated by abductive evidence.