ALBERT

Notizen: Common sense sometimes predicts events to be likely or unlikely rather than merely possible. We extend methods of qualitative reasoning to predict the relative likelihoods of possible qualitative behaviors by viewing the dynamics of a system as a Markov chain over its transition graph. This involves adding qualitative or quantitative estimates of transition probabilities to each of the transitions and applying the standard theory of Markov chains to distinguish persistent states from transient states and to calculate recurrence times, settling times, and probabilities for ending up in each state. Much of the analysis depends solely on qualitative estimates of transition probabilities, which follow directly from theoretical considerations and which lead to qualitative predictions about entire classes of systems. Quantitative estimates for specific systems are derived empirically and lead to qualitative and quantitative conclusions, most of which are insensitive to small perturbations in the estimated transition probabilities. The algorithms are straightforward and efficient.

Notizen: When Horn clause theories are combined with integrity constraints to produce potentially refutable theories, Seki and Takeuchi have shown how crucial literals can be used to discriminate two mutually incompatible theories. A literal is crucial with respect to two theories if only one of the two theories supports the derivation of that literal. In other words, actually determining the truth value of the crucial literal will refute one of the two incompatible theories.This paper presents an integration of the idea of crucial literal with Theorist, a logic-based system for hypothetical reasoning. Theorist is a goal-directed nonmonotonic reasoning system that classifies logical formulas as possible hypotheses, facts, and observations. As Theorist uses full clausal logic, it does not require Seki and Takeuchi's notion of integrity constraint to define refutable theories. In attempting to deduce observation sentences, Theorist identifies instances of possible hypotheses as nomological explanations: consistent sets of hypothesis instances required to deduce observations. As multiple and mutually incompatible explanations are possible, the notion of crucial literal provides the basis for proposing experiments that distinguish competing explanations.We attempt to make three contributions. First, we adapt Seki and Takeuchi's method for Theorist. To do so, we incrementally use crucial literals as experiments, whose results are used to reduce the total number of explanations generated for a given set of observations. Next, we specify an extension which incrementally constructs a table of all possible crucial literals for any pair of theories. This extension is more efficient and provides the user with greater opportunity to conduct experiments to eliminate falsifiable theories. A prototype is implemented in CProlog, and several examples of diagnosis are considered to show its empirical efficiency. Finally, we point out that assumption-based truth maintenance systems (ATMS), as used in the multiple fault diagnosis system of de Kleer and Williams, are interesting special cases of this more general method of distinguishing explanatory theories.

Notizen: When trying to understand a speaker's argument, it is necessary to determine what her claim is and what evidence she provides for it. It is necessary, therefore, to be able to recognize evidence relations in terms of the speaker's beliefs. This paper describes an implementation of an evidence oracle, which tests for evidence between statements and builds a model of the speaker based on the evidence relations found. This implementation is intended to be an advance in the development of practical discourse analysis systems, proposing a basis for verifying certain relationships between utterances. Another contribution of the work is a stratified speaker model which allows for varying levels of acceptance of beliefs attributed to the speaker. Integration of the implemented evidence oracle into a full discourse analyser is presented, together with output illustrating the analysis for several sample arguments. Some extensions of this approach for plan inference are also discussed.Lorsque l'on essaie de comprendre l'argument d'un locuteur, il importe de déterminer la nature de sa prétention et le type d'évidence qui l'accompagne. Par conséquent, il est nécessaire de pouvoir distinguer des relations d'évidence les croyances du locuteur. Cet article décrit la mise en oeuvre d'un oracle qui recherche l'évidence entre des énoncés et construit un modèle du locuteur en fonction des relations d'évidence constatées. Cette mise en oeuvre propose une base pour vérifier certaines relations entre des énoncés; elle se veut une contribution au développement d'un système pratique d'analyse du discours. Une autre contribution de cette recherche est l'élaboration d'un modèle de locuteur stratifyé qui tient compte de niveaux variables d'acceptation des croyances attributeés au locuteur. l'intégration de l'oracle d'évidence sous forme d'analyseur de discours est présentée, ainsi que des illustrations de l' analyse de plusieurs arguments types. Une extension de cette approche à l'inférence de plans est également discutée.

Notizen: The ability of a biological organism to visually track a perceptually significant feature in its environment has been argued to be an important feedback mechanism guiding locomotion. This paper analyzes the constraints available from the visual motion stimuli in the context of tracking. Our aim is to show that the act of tracking simplifies the decoding of egomotion parameters from motion stimuli. The constraints obtainable under tracking are utilized to analyze a possible egomotion decoding strategy for a binocular robot eye system, modeled after the human ocular tracking (smooth pursuit) mechanism. The main result of the paper is in the derivation of a closed-form solution of the egomotion parameters using feedback information concerning the movement of the tracking motors over time. The theoretical results are verified by experiments. We believe that the active tracking approach presented here is a more simple, practical, and manageable technique in a robot navigation setting, compared to passive methods.

Notizen: This paper considers computer programs that purport to do inference. It finds the essence of their reasoning ability to reside in a particular relationship which exists between the computations described by the texts of those programs and machine-independent formal representations of the theories in which they are said to reason. The principal novelties in the treatment of this question are the use of explicit “abstraction functions” between states of a computation and sets of sentences of a theory, and the notion of global “inference relations” which characterize the set of all inferences of which a system is capable. The central result of the paper is an “arrow-theoretic” characterization of the concept of a knowledge representation, along with the related concept of a knowledge representation realization. The notions of equivalence and subsumption for knowledge representations are defined, as well as equivalence of knowledge representation realizations. Finally, important issues of expression evaluation and control are addressed. The paper concludes with a slogan: “The AI is in the arrows.”Cet article traite des programmes informatiques qui prétendent faire de ľinférence. Ľauteur constate que ľessentiel de leur capacityé de raisonnement réside dans la relation particulière qui existe entre les calculs décrits par les textes de ces programmes et les représentations formelles non liées à un type de machine des théories dans le cadre desquelles ils sont supposés raisonner. Les principales nouveautés dans le traitement de cette question sont ľutilisation de fonctions ?abstraction explicites éntre les etats ?un calcul et les ensembles de phrases ?une théorie, et la notion de relations ?inférence globales qui caractérisent ľensemble de toutes les inférences dont un système est capable. Le principal résultat de cet article est une caractérisation du concept de la représentation des connaissances, ainsi que du concept connexe de la réalisation de la représentation des connaissances. Les notions ?équivalence et de sous-somption pour la représentation des connaissances sont définies, ainsi que ľéquivalence des réalisations de la représentation des connaissances. Enfin, des questions importantes comme ľanalyse et le contrǒle ?expressions sont traitées.

Notizen: Abstract: The rapid growth in the use of expert systems has led to widespread interest in techniques for their evaluation, which includes verification and validation. Evaluation techniques in the software engineering community are well established but there are different opinions about the effectiveness of transferring software engineering methodologies to expert systems. This paper outlines some of the issues involved in evaluating expert systems and cites almost 200 significant papers on the topic. We present it with the expectation that it will help both new and established researchers become acquainted with the literature of an important and growing field.

Notizen: Abstract: This paper describes some of the problems encountered in developing a user interface for a commercial expert system. The emphasis is on the practical side of interface implementation while attempting to follow the many guidelines presented in the literature.

Notizen: Abstract: The need to integrate knowledge-based systems (KBS) with mainstream data processing (DP) systems is increasingly recognised. It is also recognised that there is no ideal mode of KBS integration: different types of problem call for different modes of integration. Many large organisations are now facing a serious problem in the operation of their existing DP systems. This paper focuses on the problem of providing intelligent support to the (often inexperienced) terminal users of complex mainframe systems—but without initiating any significant changes to the existing DP environment. Over the past few years we have been attempting to evolve a solution to this type of problem that is satisfactory from both a human and a software engineering perspective. This paper outlines the previous work and considerations underlying our current approach. Starting from a purely mainframe-based mode of integration, this has entailed locating an increasing number of an integrated KBS-DP system's components onto a front-end workstation. In our current solution, both user interface and KBS modules reside on a workstation linked to the (unchanged) mainframe environment. The key element of this approach is terminal operator emulation: a KBS-controlled 'script’ facility on a workstation which emulates the flexible keyboard behaviour of an experienced mainframe terminal user interacting with an existing DP system. The software tools and techniques employed in implementing this solution are discussed and details of a large operational system are presented as an example application.

Notizen: Abstract: The aim of this paper is to describe a system whereby an intensive care unit patient's condition can be monitored and assessed over time using mostly blood gas analysis data. The first six sections analyse some major temporal reasoning approaches (with limited examples from other related areas) with the aim of selecting one for the task at hand. The last sections describe the PONJIP temporal blood gas analysis expert system with its main components, including user interfaces customised to the doctor's requirements. Since the overall aim of this paper is to describe an expert system used for the evaluation of the possible contribution of temporal medical reasoning from blood gas data, research specific to other application areas or data will not be covered here.

Notizen: Abstract: Solving the customer's LAN/WAN interconnect problem is difficult because of the need to explore many possible configurations (e.g. bridging/routing, packet/channels) and then to choose the best configuration using economic, performance and other criteria. The rapid introduction of new standards, protocols and products to the networking field brings additional complications to the solution and can cause confusion when configuring a network. ALCA is intended to generate all feasible LAN/WAN configuration possibilities automatically and to pick the most appropriate solution to solve the customer's problem, while specifically addressing open systems interconnection (OSI) standards. Matching communications protocols while searching all possible configurations is notoriously slow even on a computer. We show how the search speed can be significantly improved by using expert system knowledge compilation, a computer-aided software engineering (CASE) technique. ALCA is based on a centrally updated knowledge base of various local area networking products and their interconnect possibilities. ALCA also allows querying to find out protocol interfaces supported by a particular product/service. Finally, it includes a graphic user interface and context-sensitive menus to reduce user information load. ALCA is intended to be used by the field personnel involved in pre-sales support, by the data communication product managers, and as an educational tool for novice communication product managers.

Notizen: Abstract: Artificial intelligence has emerged from the toy problem world and is being applied to real domains in a more general way, the existence of several large application systems supporting the belief that a generation of smarter and more general systems will be developed. However, a new problem, sometimes referred to as the fusion problem, has been identified, which acts to restrict the development of such systems. This paper explains the nature of the problem, and by examining a proposed expert system in economics (ESE), discussing three approaches to a prototype ESE and the problems associated with them, draws some conclusions with regard to data fusion and expert co-operation.

Notizen: Abstract: This paper presents work undertaken as part of a project concerned with the development of a fully automated industrial radiographic inspection system, based on both conventional image-processing techniques for the detection and analysis of defects in the radiographic image, and intelligent knowledge-based (1KB) techniques for the classification and evaluation of defect data against the quality assurance requirements of the inspection process. In this paper the 1KB defect classification system is presented. This system is based on a hierarchical frame-based knowledge representation and a backward-chaining production rule system. Examples of the frame structures, frame taxonomies and the data-driven procedures, which maintain the knowledge base are given, along with an outline of the defect classification rules and the inference mechanism for dealing with uncertainty by means of confidence factors.

Notizen: Abstract: The excessive time devoted to the development, testing and maintenance of expert systems (ESs) needs to be reduced radically. What is needed is general-purpose software that can interlace the extracting, structuring, testing, and encoding of knowledge gained from debriefing the expert in any field. LAPS (Logic Aids for Problem Solving) is just such a program and has been successfully used to produce an ES in the domain of submarine diving officers. This automatically-encoded ES was produced by each expert entering his responses to the LAPS series of strategic-induction, AI-jargon-free queries. Currently, the LAPS prototype has four core sessions or functions: (1) the initial or sample-solution session; (2) the dechunking or hidden-knowledge session; (3) the alternatives or completeness-testing session; and (4) the automatic rule-production session. After any session the interviewee can use the fourth function of LAPS to produce frame-laden rules (in M.I, for now; later, in CLIPS). These rules can of course be used to carry out consultation sessions, though fewer of these are needed. LAPS has itself been translated from M.I to C to provide a very fast program on an almost universally available platform — the microcomputer. In accord with the overall technical approach to this project, other successful offline rigorous interviewing functions are being added to LAPS. Other code additions pertain to (a) graphical and other user-friendly reasoning-aid enhancements; (b) a series of simpler follow-up queries; and (c) the fleshing out of many stubs, or functions having intermediate results but as yet no processing.

Notizen: Abstract: This paper describes a shell for cooperating expert systems that has been developed at the University of Porto. The main goal of this shell is two-fold: to generate a community of cooperative knowledge-based systems and to develop several special reasoning techniques which can be used under a distributed and cooperative paradigm. UPShell is able to convert a set of generated intelligent systems (ISs) into a community of cooperative ISs. In this first version it is already possible to generate different intelligent systems which are able to run ‘simultaneously’ as separate Unix processes and, using a message-passing mechanism, to communicate among themselves. They can be set to pursue an overall goal in a cooperative way. Moreover, several tasks can be given to each IS to be solved simultaneously, and the IS can switch from task to task according to dynamic priorities reflecting the urgency attached to the specific sub-tasks that emerge. The shell described here may also be used to test, within a distributed environment, some time-bounded reasoning techniques that are presently being developed. The paper has three main parts: a general overview of the UPShell (Section 1); a tutorial explaining, by means of examples, how to use the package (Section 2); and, finally, some considerations on the reasoning techniques used and future improvements (Sections 3–5).

Notizen: Book reviewd in this article:Expert Systems in Production and Services II: From Assessment to Action?Thomas Bernold, Ulrich Hillenkamp (Eds.)Expert Systems: Human Issues. D. Berry, A. Hart.Prolog and its Applications: A Japanese Perspective. Fumio Mizuguchi (Ed.)

Notizen: Abstract: This article describes the integration of an expert system shell (AllRound Expert System, AR/XPS) with a Unix-based 4GL application generator (Task Flow Management, TFM). TFM supports the client-server architecture of distributed applications and provides access to several relational databases, including Informix and Oracle. A short overview of the major features of TFM is given, with special regard to characteristics relevant for the integration of an expert system. The main focus lies in the architecture of AR/XPS, which contains rule-based and frame-based knowledge representation methods and which was implemented in Prolog; the user interface is totally unified with that of TFM, so that the user sees only one system and need not be aware that an expert system is present. AR/XPS has access to all features of TFM and to any databases present, so that an expert system can take decisions according to the current status of the database. At the end of this article a description of the Knowledge Base Creator is given and some remarks are made regarding our experience with pilot applications.

Notizen: Abstract: Instigating the selection of a methodological approach for the development of a knowledge-based system (KBS) should not be trivialised. This paper sets out to enlighten the reader about the needs and difficulties in comparing methodologies. In doing so, a descriptive framework is presented which allows the depiction of essential features of a methodology. This is intended to simplify the comparative study of KBSs that a knowledge engineer might be involved in.

Notizen: ProKappa is an expert system tool kit developed by IntelliCorp in C using X-Windows for graphical interface. The SUN 3 or 4 version costs $9995/£7000 and $2595/£1800 for the interactive C environment. This review covers ProKappa version 1.1, installed on a colour SUN SPARC Station 1 with the minimum recommended values of 12 Mb of memory and 30 Mb of swap space. IntelliCorp can be contacted in the USA at 1975 El Camino Real West, Mountain View, CA 94040-2216 (+1 (415) 965-5500) and in the UK at 10 Jewry Street, Winchester, Hampshire SO23 8RZ (+44 (0)962 735348).

Notizen: Book reviewed in this article:Integrating knowledge-based and database management systems, Ruth KerryNonmonotonic reasoning: logical foundations of commonsense Gerhard BrewkaBuilding knowledge-based systems: towards a methodology John S. EdwardsThe Rise of the Expert Company—How Visionary Companies are Using Artificial ntellig ence to Achieve Higher Productivity and Prof its, Edward Feigenbaum

Notizen: Abstract: The purpose of this paper is to draw attention to some important barriers to the practical application of expert systems to management problems. By drawing attention to these barriers, we show two findings: (a) that certain types of knowledge found in management situations are more likely to lead to the successful adoption of management expert systems (MESs) than others; (b) that a potential exists for an alternative form of MES application. The first finding has implications for knowledge elicitation. We propose that MESs developed from academic knowledge, which we define in specific terms, are more likely to succeed than those developed from working with managers in the field. Alternatively, feasibility studies should include an analysis of the type of knowledge which a manager is capable of providing to ensure that suitable knowledge is available for knowledge elicitation purposes. The second finding has implications for the developers of MESs at both research and commercial levels. We suggest the need for a check-list MES application, capable of converting responses to strategic and operational questions into managerial actions.

Notizen: Artificial Intelligence Terminology: A Reference Guide Colin BeardonExpert Systems: A Decision Support Approach - With Applications in Management and Finance M. Klein and L. B. Methlie, AddisonThe Logical Basis for Computer Programming, Vol. 2: Deductive Systems Zohar Manna and RichardProlog Programming for Artificial Intelligence Ivan Bratko

Notizen: Prolog III is available from PrologIA Luminy, Case 919 13288, Marseille Cedex 09, France. Tel: +33 91 26 86 36. Prices (French francs): Sun 3, 70 000; Sparc 100 000; Mac Plus/SE 30 000; Mac SE 30/II38 000; IBM PC 386 38 000. Prices exclude VAT. Universities are entitled to a 60% discount and research centres to 30%. Systems are also available for HP9000 (series 300 and series 800), VAX and DECstation.

Notizen: Abstract: In this article we describe how two popular AI representation techniques—frames and production systems— can be usefully combined under a general AI object-oriented approach to problems which arise in the domain of hardware fault diagnosis. One of the main advantages of such a combination is that causal reasoning— crucial to the domain under consideration — can also be naturally and effectively represented. In order to put the combination of frames and production systems on a sound methodological footing, we first provide a knowledge engineering methodology which is an object-oriented re-interpretation of ontological analysis.

Notizen: Book reviewd in this article:Developing Expert Systems A Knowledge Engineer's Handbook for Rules and Objects, Edmund C. Payne, Robert C. McArthurBuilding Large Knowledge-Based Systems. Douglas B. Lenat, R. V. Guha.Designing Systems A Guide to Selecting Implementation Techniques. Paul J. Kline and Steven B. Dolins.

Notizen: In this paper we provide an implementation strategy to map a functional specification of an utterance into a syntactically well-formed sentence. We do this by integrating the functional and the syntactic perspectives on language, which we take to be exemplified by systemic grammars and tree adjoining grammars (TAGs) respectively. From systemic grammars we borrow the use of networks of choices to classify the set of possible constructions. The choices expressed in an input are mapped by our generator to a syntactic structure as defined by a TAG. We argue that the TAG structures can be appropriate structural units of realization in an implementation of a generator based on systemic grammar and also that a systemic grammar provides an effective means of deciding between various syntactic possibilities expressed in a TAG grammar. We have developed a generation strategy which takes advantage of what both paradigms offer to generation, without compromising either.

Notizen: Tree-adjoining grammars (TAG) have been proposed as a formalism for generation based on the intuition that the extended domain of syntactic locality that TAGs provide should aid in localizing semantic dependencies as well, in turn serving as an aid to generation from semantic representations. We demonstrate that this intuition can be made concrete by using the formalism of synchronous tree-adjoining grammars. The use of synchronous TAGs for generation provides solutions to several problems with previous approaches to TAG generation. Furthermore, the semantic monotonicity requirement previously advocated for generation grammars as a computational aid is seen to be an inherent property of synchronous TAGs.

Notizen: A large number of wording choices naturally occurring in English sentences cannot be accounted for on semantic or syntactic grounds. They represent arbitrary word usages and are termed collocations. In this paper, we show how collocations can enhance the task of lexical selection in language generation. Previous language generation systems were not able to account for collocations for two reasons: they did not have the lexical information in compiled form and the lexicon formalisms available were not able to handle the variations in collocational knowledge. We describe an implemented generator, Cook, which uses a wide range of collocations to produce sentences in the stock market domain. Cook uses a flexible lexicon containing a range of collocations, from idiomatic phrases to word pairs that were compiled automatically from text corpora using a lexicographic tool, Xtract. We show how Cook is able to merge collocations of various types to produce a wide variety of sentences.

Notizen: Natural language generation systems should choose nouns by searching for lexical units that (i) are known to the user; (ii) truthfully describe the object being lexicalised; (iii) convey sufficient information to fulfill the system's underlying communicative goals; and (iv) are maximal under a lexical preference function. This model of lexical choice allows a clean separation to be made between what the system knows about the object being lexicalized and what it wishes to communicate about this object. The model also allows lexical choice to be biased towards basic-level and other preferred lexical units.Les systèmes de génération du langage naturel devraient choisir des noms en recherchant des unités lexicales (i) qui soient connues de ľutilisateur, (ii) qui décrivent fidèlement ľobjet soumis à un traitement lexical, (iii) qui transmettent suffisamment ?information pour répondre aux objectifs de communication du système, et (iv) qui soient maximales dans une fonction de préférence lexicale. Ce modéle de choix lexical permet une nette séparation entre ce que le système sait à propos de ľobjet soumis à un traitement lexical et ce qu'il souhaite communiquer à propos de cet objet. Le modèle permet également au choix lexical ?avoir un préjugé vis-à-vis les unités lexicales de niveau de base et autres unités lexicales préférées.

Notizen: In this paper, we describe a general framework for the generation of referring expressions. Given an intended referent in the domain, domain-dependent mapping rules build a domain-independent level of recoverable semantic structure, which encodes the semantic content to be recovered by the hearer; a second set of mapping rules then builds an abstract syntactic structure, which is unified with a grammar and lexicon to produce a surface noun phrase. Within this framework, (0 we present a detailed description of the algorithms required to build the recoverable semantic structure, and discuss how the problem of infinite recursion can be handled when using relations in descriptions; and (/;') we show how the two levels of representation facilitate the generation of one-anaphoric noun phrases.

Notizen: Previous work in natural language génération has exploited discourse focus to guide the selection of propositional content and the génération of referring expressions (e.g., pronominalization, définite noun phrase génération). However, there are many other sources of contextual information which can be used to constrain linguistic realization. The realization of certain classes of temporal and spatial référents, for example, can be guided by more detailed models of time and space. Therefore, this article first identifies a number of contextual coordinates or points of référence (known as indexicals in linguistics). Next, we formalize three of these contextual coordinates – topic, time, and space – in a computational model. In doing so, the article describes the use of a Reichenbachian temporal model which is exploited to guide the realization of verb tense and aspect as well as the realization of temporal référents (e.g., temporal connectives and adverbials such as “meanwhile” and “ten minutes later”). The article then describes a spatial model which is used to guide the linguistic realization of spatial référents (e.g., “here,”“there”) and spatial adverbials (e.g., “two miles away”). We discuss the relation between these temporal and spatial models and illustrate their use in generating text from two different application systems. We find that the combined use of topical, temporal, and spatial contextual coordinates enhances the fluency, connectivity, and conciseness of the resulting text.

Notizen: Naturally occurring discourse often contains expressions such as “however,”“as I have stated before,” and “next.” These expressions, denoted meta-comments, carry important information which helps a hearer speed up the comprehension process. In this paper, we present a mechanism for the generation of meta-comments and their incorporation into computer-generated discourse. This mechanism is based on a simple model of the anticipated effect of the messages to be conveyed and of the meta-comments on the hearer's comprehension. It was implemented in a system called FIGMENT, which generates commentaries on the solution of algebraic equations.Le discours naturel contient souvent des expressions comme 〈〈 cependant 〉〉 et 〈〈 comme je ľai dit auparavant 〉〉. Ces expressions, appelées méta-commentaires, contiennent des informations importantes qui aident à accélérer le processus de compréhension de ľ auditeur. Dans cet article, ľ auteur présente un mécanisme de génération de métacommentaires et ? intégration de ceux-ci au discours créé par ordinateur. Ce mécanisme est basé sur un modèle simple de ľeffet anticipé des messages devant ětre transmis et de ľ';effet des méta-commentaires sur la compréhension de ľ auditeur. II a été mis en oeuvre dans un système appele FIGMENT, qui génère des commentaires sur la solution ?équations algébriques.

Notizen: This paper analyzes how a spectrum of architectural and structural ideas fit together to provide the required functionality for explanation generation. Several information processing tasks involved in the choice and organization of the content of an explanation are identified. These are best modeled by distinct mechanisms; hence a particular class of hybrid planning architectures most clearly reflects the nature of the explanation task. The architecture is exemplified by a description of an implemented explanation planner. Various implications of the architecture are discussed, including a classification of structuring relations based on their sources and roles in planning; the elimination of goal-posting preconditions from goal refinement operators; and the level at which nondeterminism is handled.

Notizen: Much research in text planning has been oriented towards identifying the rhetorical component of communication knowledge that allows for the structuring of coherent multi-sentential or even multi-paragraph text. However, several researchers have recently found rhetoric-based top-down approaches to be inadequate for their particular needs. In this paper, we explore the issue by introducing the concept of “domain communication knowledge” as a new analytical tool. By analyzing two existing rhetoric-based text planners, and by closely analyzing a particular genre, namely reports, we conclude that domain-specific knowledge about communication is pervasive and should be represented explicitly for the text planning task.La plupart des travaux de recherche dans la planification de textes ont été orientés vers ľ identification de la composante rhétorique de la connaissance de la communication, qui permet de structurer de manière cohérente un texte à phrases ou à paragraphes multiples. ?autre part, plusieurs chercheurs ont récemment constaté que les approches descendantes basées sur la rhétorique ne convenaient pas à leurs besoins particuliers. Dans cet article, les auteurs examinent la question en utilisant un nouvel outil analytique, soit le concept de 〈〈 connaissance du domaine reliée à la communication 〉〉 En étudiant deux planificateurs de textes existants et an analysant de très près un genre particulier, soit des rapports, les auteurs concluent que la connaissance spécifique du domaine à propos de la communication est subtile et devrait faire ľobjet ?une representation explicite pour la planification de textes.

Notizen: Text planning is traditionally done as a separate process, independent of linguistic realization. However, I show in this paper that it is essential for a text planner to know the linguistic consequences of its actions. The choice of how to realize an element affects how much information is conveyed (e.g., “Floyd arrived safely in Boston” vs. “Floyd landed safely at Logan”) and what other information can be added (e.g., “Peter was deciding for an hour” vs. *“Peter made a decision for an hour”). The choice of realization also affects how the relative salience of the elements being expressed will be perceived (e.g., “The green car is in the garage” vs. “The car in the garage is green”). I have defined an intermediate level of representation for text planning, called the Text Structure. It is an abstract linguistic level that reflects germane linguistic constraints while abstracting away from syntactic detail. This representation allows the text planner to have greater control over the decisions, so that it can take advantage of the expressiveness of language to convey subtleties of meaning. More importantly, the Text Structure allows the generation process overall to be incremental, since it ensures that the text plan being composed will always be expressible in the language.La planification de textes est habituellement réalisée séparément, sans tenir compte de la réalisation linguistique. Cependant, ľauteur demontre dans cet article qu'il est essentiel à un planificateur de textes de connai̊tre les conséquences linguistiques de ses actions. Le choix de la méhode de réalisation ?un élément influe sur la quantité?informations qui est transmise (par ex.: 〈〈 Floyd est arrivé sain et sauf à Boston 〉〉 et 〈〈 Floyd a atterri sain et sauf àľaéroport Logan 〉〉) et sur quelle autre information peut ětre ajoutée. Le choix de la méthode de réalisation influe également sur la façon dont les caractères saillants des ééments qui sont exprimés seront perçus (par ex.: 〈〈 la voiture verte est dans le garage 〉〉 et 〈〈 la voiture dans le garage est verte 〉〉). L'auteur a défini un niveau intérmediaire de représentation pour la planification de textes qu'il a appelé structure de texte. Il; s'agit ?'un niveau linguistique abstrait qui refléte les contraintes linguistiques appropriées tout en s'éloignant du détail syntaxique. Cette représentation permet au planificateur de textes ?avoir un meilleur contrǒle des décisions, et done de tirer profit de la force ?expression du langage afin de tenir compte des subtiliés de sens. Plus important encore, la structure de texte permet au processus de génération ?ětre incrémentiel, car elle s'assure que le plan de texte en voie de composition soit toujours exprimable dans le langage.

Notizen: This paper notes some structural features of naturally occurring, extended explanations and argues that current generation methodologies are inadequate for determining high-level structure. We then present a computational model for generating high-level structure, which is based on the hypothesis that high-level structure is distinct from text structure at lower levels and is determined by the material to be conveyed rather than by independent, recursive text structuring devices. We propose that high-level structure can be determined by bottom-up processes that attempt to satisfy speaker, listener, and compositional goals, after which top-down strategies can be used to organize the material about the selected framework.Cet article relève certaines caractéristiques structurales ?explications étendues à occurrence naturelle et soutient que les méthodes de génération courantes ne permettent pas de déterminer adéquatement une stratégie de haut niveau. Les auteurs présentent ensuite un modèle computationnel de géneration de structures de haut niveau qui est basé sur ľhypothèse qu'une structure de haut niveau est différente ?une structure de texte aux nivaux inféiieurs, et qu'elle est déterminée par les données qui doivent ětre transmises plutǒt que par des mécanismes indépendants et récursifs de structure de texte. lis soutiennent qu'une structure de haut niveau peut ětre déterminée àľaide ?'un processus ascendant qui essaie de rencontrer les objectifs du locuteur, de ľauditeur et de composition, et que des stratégies descendantes peuvent ensuite ětre utilisées pour organiser les données a propos du cadre choisi.

Notizen: In this paper, we analyze the logical definitions of model-based diagnosis recently presented in the literature, and we propose a unified framework (based on the integration of abductive and consistency-based reasoning) in which most of such definitions can be captured. This allows us to single out the existence of a spectrum of alternatives in the logical definition of diagnosis. A lot of attention in the paper is devoted to analyzing the differences among the definitions in the spectrum. In particular, we show that the definitions can be compared on the basis of their restrictive-ness and we relate such a restrictiveness with the completeness of the model of the system to be diagnosed.Dans cet article, les auteurs analysent les définitions logiques de diagnostics basés sur un modèle dont il a été question récemment dans certains ouvrages. Ils proposent un cadre unifié (basé sur l'intégration du raisonnement abductif et du raisonnement basé sur la consistance) à l'interieur duquel la plupart de ces définitions peuvent ětre regroupées. Cette particularité permet de mettre en lumière l'existence d'un spectre d'alternatives dans la définition logique du diagnostic. Cet article accorde une attention toute particulière à l'analyse des différences entre les définitions du spectre. En outre, les auteurs démontrent que les définitions peuvent ětre compareées en fonction de leur caractère restrictif, qui est ensuite mis en relation avec la complétude du modèle du système faisant l'objet d'un diagnostic.

Notizen: The use of multivalued logics for knowledge representation and nonmonotonic reasoning has often been advocated, in particular within the general framework proposed by Ginsberg in his paper “Multivalued logics: a uniform approach to reasoning in artificial intelligence.” His system is based on a multivalued logic with an arbitrary number of truth values classified with respect to two partial orders, a truth order and a knowledge order. This classification is very interesting and gives an intuitive appeal to the framework. In this paper the work by Ginsberg is critically reviewed, pointing out some flaws and ways to overcome them. Moreover, we present some ideas on how to modify the original schema in order to obtain a more semantically well-founded framework.L'utilisation de la Iogique multivalente pour la représentation des connaissances et le raisonnement non monotone a souvent été préconisée, en particulier à l'intérieur du cadre général proposé par Ginsberg dans son article intitulé? Multivalued logics: a uniform approach to reasoning in artificial intelligence ? Son système est basé sur une logique multivalente comportant un nombre arbitraire de valeurs de vérité classées selon deux ordres partiels: un ordre de vérité et un ordre de connaissances. Cette classification est très intéressante et donne un attrait intuitif au cadre. Dans cet article, l'auteur examine le travail de Ginsberg, y relève des lacunes et propose des moyens de les corriger. De plus, il expose certaines idées en vue de modifier le schéma original et ainsi obtenir un meilleur cadre du point de vue de la sémantique.

Notizen: The relationship between Reiter's default logic and general logic programs, which may contain negative subgoals in rule bodies, has been discussed in the literature by translating logic programs to default logic theories. Here, we present a method to translate some default logic theories to general logic programs, and study the extensions of default logic theories with the stable model semantics of logic programming. Based on the translation method, we show that another semantics of logic programming, the well-founded semantics, can be used to define a new version of default logic, which is more cautious than the original one. This enables the existing proof procedures for the well-founded semantics to perform default inference. We also study the property of cumulative monotonicity for both default logic theories and general logic programs under the two different semantics. As a direct application of the translation method, logic programs can be used to make default inference for semantic networks with exceptions.La relation entre la logique par défaut de Reiter et les programmes logiques généraux, qui peuvent comporter des sous-buts négatifs dans les ensembles de règies, a déjàété discutée en traduisant les programmes logiques en théories de la logique par défaut. Dans cet article, les auteurs proposent une méthode pour traduire certaines théories de la logique par defaut en programmes logiques généraux et étudient les extensions des theories de la logique par defaut avec la sémantique de modéle stable de la programmation logique. En se basant sur la méthode de traduction, les auteurs démontrent qu'une autre sémantique de la programmation logique, la sémantique bien fondée, peut ětre utilisée pour définir une nouvelle version de la logique par défaut, qui est plus prudente que la première. Cela permet aux procédures de preuve existantes de la sémantique bien fondée d' effectuer l' inférence par défaut. Les auteurs examinent également les caracteristiques de la monotonicité cumulative pour les theories de la logique par défaut et les programmes logiques généraux selon les deux différentes sémantiques. Une application directe de la méthode de traduction est la possibilityé d' utiliser les programmes logiques pour effectuer de l' inférence par défaut pour les réseaux sémantiques avec exceptions.

Notizen: The development of efficient algorithms for learning from large relational databases is an important task in applicative machine learning. In this paper, we study knowledge discovery in relational databases and develop an attribute-oriented learning method which extracts generalization rules from relational databases. The method adopts the artificial intelligence “learning-from-examples” paradigm and applies in the learning process an attribute-oriented concept tree ascending technique which integrates database operations with the learning process and provides a simple and efficient way of learning from databases. The method learns both characteristic rules and classification rules of a learning concept, where a characteristic rule characterizes the properties shared by all the facts of the class being learned; while a classification rule characterizes the properties that distinguish the class being learned from other classes. The learning result could be a conjunctive rule or a rule with a small number of disjuncts. Moreover, learning can be performed with databases containing noisy data and exceptional cases using database statistics. Our analysis of the algorithms shows that attribute-oriented induction substantially reduces the computational complexity of the database learning process.Le développement d'algorithmes efficaces permettant l'apprentissage à partir de bases de donnees relationnelles est une fonction importante de l'apprentissage automatique applicatif. Dans cet article, les auteurs examinent la découverte des connaissances dans les bases de données relationnelles et élaborent une méthode d'apprentissage orientée sur l'attribut qui extrait des bases de données relationnelles les règies de généralisation. La méthode adopte le paradigme d'apprentissage à partir d'exemples et applique au processus d'apprentissage la technique de l'arbre des concepts orientés sur l'attribut qui incorpore les opérations de base de données au processus d'apprentissage, ce qui permet d'obtenir une méthode simple et efficace d'apprentissage à partir des bases de données. La méthode fait l'apprentissage des règies caractéristiques et des règies de classification d'un concept d'apprentissage; la règie caractéristique qualifie les pro-priétés communes à tous les faits d'une categorie faisant l'objet d'un apprentissage alors que la règie de classification caractérise les propriétés qui distinguent la catégorie faisant l'objet d'un apprentissage des autres catégories. Le résultat peut ětre une règie conjonctive ou une règie ayant un petit nombre de disjonctifs. Qui plus est, 1′apprentissage peut se faire avec des bases de données contenant des donnees bruitees et des cas exceptionnels utilisant des statistiques de bases de données. L'analyse des algorithmes démontre que l'induction orientée sur l'attribut réduit considérablement la complexité informàtique du processus d'apprentissage des bases de données.

Notizen: Self-knowledge is a concept that is present in several philosophies. In this article, we consider the issue of whether or not a learning algorithm can in some sense possess self-knowledge. The question is answered affirmatively. Self-learning inductive inference algorithms are taken to be those that learn programs for their own algorithms, in addition to other functions.La connaissance de soi est un concept qui se retrouve dans plusieurs philosophies. Dans cet article, les auteurs s'interrogent à savoir si un algorithme d' apprentissage peut dans une certaine mesure posséder la connaissance de soi. lis apportent une reponse positive a cette question. Les algorithmes d'inference inductive autodidactes sont ceux qui font l'apprentissage de programmes pour leurs propres algorithmes, en plus d' autres fonctions.

Notizen: This article describes a polynomial-time, O(n3), planning algorithm for a limited class of planning problems. Compared to previous work on complexity of algorithms for knowledge-based or logic-based planning, our algorithm achieves computational tractability, but at the expense of only applying to a significantly more limited class of problems. Our algorithm is proven correct, and it always returns a parallel minimal plan if there is a plan at all.Cet article décrit un algorithme de planification de temps polynomial O(n3) pour une classe restreinte de problemes de planification. Contrairement aux travaux précédents sur la complexité des algorithmes pour la planification basée sur la logique ou les connaissances, l' algorithme dont il est question dans cet article permet d' obtenir la tractabilityé computationnelle; cependant, il ne peut ětre appliqué qu'à une catégorie beaucoup plus restreinte de problèmes. Cet algorithme s'est done révélé correct et il génère toujours un plan minimal en parallèle lorsqu'il y en a un.

Notizen: A question-answering program that provides access to a large amount of data will be most useful if it can tailor its answers to each individual user. In particular, a user's level of knowledge about the domain of discourse is an important factor in this tailoring if the answer provided is to be both informative and understandable to the user. In this research, we address the issue of how the user's domain knowledge, or the level of expertise, might affect an answer. We present TAILOR, a flexible computer system that takes into account this knowledge to provide an answer that is appropriate for users with varying levels of expertise (including novices and experts), without requiring an a priori set of user types.Un programme qui donne accès à un système de connaissances à de nombreux et divers utilisateurs serait plus utile s'il pouvait adapter ses réponses pour chaque utilisateur. En particulier, le savoir de l'utilisateur est un facteur important pour que la réponse soit compréhensible et informative. Dans cet article, nous étudions le rǒle de la connaissance de l'utilisateur pour former un texte qui soit approprié pour la personne qui le lira. Nous présentons TAILOR, un système de génération flexible qui est sensitif au niveau de connaissance de l'utilisateur. TAILOR peut générer des textes appropries pour des utilisateurs aux niveaux de connaissances très variés, sans avoir besoin ?un groupe de stéréotypes définis a priori.

Notizen: We introduce a new probabilistic approach to dealing with uncertainty, based on the observation that probability theory does not require that every event be assigned a probability. For a nonmeasurable event (one to which we do not assign a probability), we can talk about only the inner measure and outer measure of the event. In addition to removing the requirement that every event be assigned a probability, our approach circumvents other criticisms of probability-based approaches to uncertainty. For example, the measure of belief in an event turns out to be represented by an interval (defined by the inner and outer measures), rather than by a single number. Further, this approach allows us to assign a belief (inner measure) to an event E without committing to a belief about its negation -E (since the inner measure of an event plus the inner measure of its negation is not necessarily one). Interestingly enough, inner measures induced by probability measures turn out to correspond in a precise sense to Dempster-Shafer belief functions. Hence, in addition to providing promising new conceptual tools for dealing with uncertainty, our approach shows that a key part of the important Dempster-Shafer theory of evidence is firmly rooted in classical probability theory.Cet article présente une nouvelle approche probabiliste en ce qui concerne le traitement de l'incertitude; celle-ci est basée sur l'observation que la théorie des probabilityés n'exige pas qu'une probabilityé soit assignée à chaque événement. Dans le cas d'un événement non mesurable (un événement pour lequel on n'assigne aucune probabilityé), nous ne pouvons discuter que de la mesure intérieure et de la mesure extérieure de l'évenément. En plus d'éliminer la nécessité d'assigner une probabilityéà l'événement, cette nouvelle approche apporte une réponse aux autres critiques des approches à l'incertitude basées sur des probabilityés. Par exemple, la mesure de croyance dans un événement est représentée par un intervalle (défini par la mesure intérieure et extérieure) plutǒt que par un nombre unique. De plus, cette approche nous permet d'assigner une croyance (mesure intérieure) à un événement E sans se compromettre vers une croyance à propos de sa négation -E (puisque la mesure intérieure d'un événement et la mesure intérieure de sa négation ne sont pas nécessairement une seule et unique mesure). II est intéressant de noter que les mesures intérieures qui résultent des mesures de probabilityé correspondent d'une manière précise aux fonctions de croyance de Dempster-Shafer. En plus de constituer un nouvel outil conceptuel prometteur dans le traitement de l'incertitude, cette approche démontre qu'une partie importante de la théorie de l'évidence de Dempster-Shafer est fermement ancrée dans la theorie classique des probabilityés.

Notizen: Many AI researchers have come to be dissatisfied with approaches to their discipline based on formal logic. Various alternatives are often suggested, including probability theory. This paper investigates the intimate connection between probability theory and various logics. We show that probability theory, broadly conceived, may be used as a formal semantics for virtually any monotonic logic. Thus, rather than being seen as competing, it is more appropriate to view formal logics as very special cases of probability theory, usually special cases that are computationally more tractable than the more general theory. Thus, probability theory and logic should be seen as complementary. Viewing probability theory in this abstract way may help to shed light on various recalcitrant problems in AI.De nombreux chercheurs dans le domaine de l'intelligence artificielle manifestent une certaine insatisfaction vis-à-vis certaines approches basées sur la logique formelle. Diverses solutions sont souvent proposées, y compris la théorie des probabilityés. Cet article analyse la relation intime entre la théorie des probabilités et diverses logiques. Il est démontré que la théorie des probabilityés, conçue de manière générale, peut ětre utilisée comme une sémantique formelle pour presque toute logique monotonique. Au lieu de percevoir les logiques formelles comme étant en opposition, il est plus approprié de les considérer comme des cas trés spéciaux de la théorie des probabilityés, habituellement plus traitables au niveau calcul que la théorie plus générate. Par conséquent, la théorie des probabilityés et la logique doivent ětre percues comme des éléments complémentaires. Le fait de considérer la théorie des probabilityés d'une manière abstraite peut contribuer à la compréhension de divers problèmes ardus dans le domaine de l'intelligence artificielle.

Notizen: NIKL was the successor to KL-ONE, the first knowledge representation system to implement “term subsumption” semantics. The primary goals for NIKL were to improve the expressivity and efficiency of KL-ONE. In the process of developing and using NIKL as the primary representational vehicle for several applications, important new representational needs were identified and addressed, and issues both theoretical and practical in knowledge representation were formulated. In this paper, we present the core of NIKL by way of description, examples, and a model theoretic semantics. We describe the motivations behind the major NIKL design decisions, and discuss NIKL'S capabilities and limitations as a working representation system.

Notizen: Abstract: The TRISTAR expert system (TRIangular STrAtegic Rules for British herbaceous vegetation) is a PC-based consultancy tool designed to help bridge the gap between academic plant ecology and practical vegetation management. It is a ‘what if facility, for use by students, naturalists, conservationists, planners, administrators; indeed anyone with an interest in environmental science and management, conservation or reconstruction of herbaceous vegetation. The system is written in Leonardo 3 (Creative Logic Limited, Uxbridge, UK). Release I of TRISTAR deals with the fundamental environmental and management processes which control the composition of British herbaceous vegetation. The system takes a specification of initial steady-state vegetation, asks the user to specify an altered environmental and/or management scenario, and predicts the composition of a new steady-state vegetation in terms of plant-strategic types. The high-level approach through plant strategies makes it possible to construct a very robust rulebase despite a lack of hard information in many areas of understanding.

Notizen: Abstract: Knowledge base verification, a part of the validation process in expert system development, includes checking the knowledge base for completeness and consistency to guard against a variety of errors that can arise during the process of transferring expertise from a human expert to a computer system. Regardless of how an expert system is developed, its developers can profit from a systematic check of the knowledge base without gathering extensive data for test runs, even before the full reasoning mechanism is functioning. Until recently knowledge base verification has been largely ignored, which has led to expert systems with knowledge base errors and no safety factors for correctness. We propose a unification-based approach for verification of a knowledge base represented in the form of production rules and facts. This approach can determine conflicting, redundant, subsumed and circular rules; redundant if-conditions in rules; dead-end rules; and cycles and contradiction in rules.

Notizen: Abstract: A case study-based approach was used to develop the knowledge base of an expert system for marketing media class selection. Cases, in the context of business learning, are detailed, written descriptions of business problems and their resolution. An expert system rule set was developed for media class selection using standardised business case studies. The system developed was validated using another set of case studies. Using case studies has two knowledge elicitation advantages: (1) the knowledge source is standardised and relatively unambiguous and (2) less domain expert involvement is required.

Notizen: Abstract: This article describes an intelligent knowledge-based object-oriented process planning (IKOOPP) system for the manufacture of progressive die plates. A die assembly is designed using a variety of standardised components based on a computer-aided design (CAD) system. A feature recognition module extracts all the pertinent geometrical properties and functional attributes of each machining feature from the CAD representation models. These properties and attributes are converted into an object-oriented representation. Knowledge of the functions of the machining feature allows process planning information to be automatically deduced. Specialised tool engineering knowledge are formulated as production rules or procedures to establish the required set of cutting tools, together with the machining allowances and sequence of operations.

Notizen: Abstract: The current trend in expert system building is domain-specific, i.e. there is one expert system for each problem domain. The increased involvement of computers in the decision-making process will inevitably lead to increased demand for expert systems. Based on the current approach to expert systems building, there will be a further proliferation of domain-specific expert systems. This is because each application area produces an expert system tailored to its requirements. This manner of producing expert systems is inadequate and an increased expectation in the performance of expert systems will eventually call for a new approach to constructing them. This paper examines the growth in the use of expert systems, looks into the limitations and problems associated with present-day domain-specific expert systems and suggests a multi-domain expert system architecture as a solution to the problem of increasingly disjointed domain-specific expert systems resulting from uncontrolled proliferation.

Notizen: Abstract: This paper describes the results of a five-year research program on production systems carried out in Vienna by a large engineering group at the Alcatel-Elin research laboratory. In the course of our investigations, a real-time expert system shell, PAMELA, was developed. We describe features of PAMELA and discuss ways in which the matching algorithm can be speeded up. It was found that the single processor version is one of the most efficient forward chainers available worldwide. Since single processor enhancements have been widely explored, we concentrated our efforts on multiple processor architectures to provide further performance gain. Efficient exploitation of parallelism in production systems requires suitable algorithms for load-balancing, without simultaneously increasing organisational or communication overheads. With this in mind, we developed a transputer-based architecture which serves as a testbedfor our new parallel matching algorithm. The execution speeds of various expert systems based on PAMELA are presented.

Notizen: Prolog++, available from Logic Programming Associates Ltd, Studio 4, Victoria Patriotic Building, Trinity Road, London SW18 3SX, UK. Tel: +44 (0)81-871 2016. Runs on both Macintosh and IBM PC.