## Fachbereich Informatik

### Refine

#### Year of publication

#### Document Type

- Preprint (346) (remove)

#### Keywords

- AG-RESY (17)
- Case-Based Reasoning (16)
- RODEO (10)
- Fallbasiertes Schliessen (9)
- Case Based Reasoning (6)
- Abstraction (5)
- Fallbasiertes Schließen (5)
- Robotics (5)
- case-based problem solving (5)
- CoMo-Kit (4)

We present a convenient notation for positive/negativeADconditional equations. Theidea is to merge rules specifying the same function by using caseAD, ifAD, matchAD, and letADexpressions.Based on the presented macroADruleADconstruct, positive/negativeADconditional equational specifiADcations can be written on a higher level. A rewrite system translates the macroADruleADconstructsinto positive/negativeADconditional equations.

The Internet has fallen prey to its most successful service, the World-Wide Web. The networksdo not keep up with the demands incurred by the huge amount of Web surfers. Thus, it takeslonger and longer to obtain the information one wants to access via the World-Wide Web.Many solutions to the problem of network congestion have been developed in distributed sys-tems research in general and distributed file and database systems in particular. The introduc-tion of caching and replication strategies has proven to help in many situations and thereforethese techniques are also applied to the WWW. Although most problems and associated solu-tions are known, some circumstances are different with the Web, forcing the adaptation ofknown strategies. This paper gives an overview about these differences and about currentlydeployed, developed, and evaluated solutions.

We have developed a middleware framework for workgroup environments that can support distributed software development and a variety of other application domains requiring document management and change management for distributed projects. The framework enables hypermedia-based integration of arbitrary legacy and new information resources available via a range of protocols, not necessarily known in advance to us as the general framework developers nor even to the environment instance designers. The repositories in which such information resides may be dispersed across the Internet and/or an organizational intranet. The framework also permits a range of client models for user and tool interaction, and applies an extensible suite of collaboration services, including but not limited to multi-participant workflow and coordination, to their information retrievals and updates. That is, the framework is interposed between clients, services and repositories - thus "middleware". We explain how our framework makes it easy to realize a comprehensive collection of workgroup and workflow features we culled from a requirements survey conducted by NASA.

In recent years several computational systems and techniques fortheorem proving by analogy have been developed. The obvious prac-tical question, however, as to whether and when to use analogy hasbeen neglected badly in these developments. This paper addresses thisquestion, identifies situations where analogy is useful, and discussesthe merits of theorem proving by analogy in these situations. Theresults can be generalized to other domains.

In this report we give an overview of the development of our new Waldmeisterprover for equational theories. We elaborate a systematic stepwise design process, startingwith the inference system for unfailing Knuth - Bendix completion and ending up with animplementation which avoids the main diseases today's provers suffer from: overindulgencein time and space.Our design process is based on a logical three - level system model consisting of basicoperations for inference step execution, aggregated inference machine, and overall controlstrategy. Careful analysis of the inference system for unfailing completion has revealed thecrucial points responsible for time and space consumption. For the low level of our model,we introduce specialized data structures and algorithms speeding up the running system andcutting it down in size - both by one order of magnitude compared with standard techniques.Flexible control of the mid - level aggregation inside the resulting prover is made possible by acorresponding set of parameters. Experimental analysis shows that this flexibility is a pointof high importance. We go on with some implementation guidelines we have found valuablein the field of deduction.The resulting new prover shows that our design approach is promising. We compare oursystem's throughput with that of an established system and finally demonstrate how twovery hard problems could be solved by Waldmeister.

Die Verfahren der Induktiven Logischen Programmierung (ILP) [Mug93] haben die Aufgabe, aus einer Menge von positiven Beispielen E+, einer Menge von negativen Beispielen E und dem Hintergrundwissen B ein logisches Programm P zu lernen, das aus einer Menge von definiten Klauseln C : l0 l1, : : : ,ln besteht. Da der Hypothesenraum für Hornlogik unendlich ist, schränken viele Verfahren die Hypothesensprache auf eine endliche ein. Auch wird oft versucht, die Hypothesensprache so einzuschränken, dass nur Programme gelernt werden können, für die die Konsistenz entscheidbar ist. Eine andere Motivation, die Hypothesensprache zu beschränken, ist, dass das Wissen über das Zielprogramm, das schon vorhanden ist, ausgenutzt werden soll. So sind für bestimmte Anwendungen funktionsfreie Hypothesenklauseln ausreichend, oder es ist bekannt, dass das Zielprogramm funktional ist.

Conditional Compilation (CC) is frequently used as a variation mechanism in software product lines (SPLs). However, as a SPL evolves the variable code realized by CC erodes in the sense that it becomes overly complex and difficult to understand and maintain. As a result, the SPL productivity goes down and puts expected advantages more and more at risk. To investigate the variability erosion and keep the productivity above a sufficiently good level, in this paper we 1) investigate several erosion symptoms in an industrial SPL; 2) present a variability improvement process that includes two major improvement strategies. While one strategy is to optimize variable code within the scope of CC, the other strategy is to transition CC to a new variation mechanism called Parameterized Inclusion. Both of these two improvement strategies can be conducted automatically, and the result of CC optimization is provided. Related issues such as applicability and cost of the improvement are also discussed.

We present a distributed system, Dott, for approximately solving the Trav-eling Salesman Problem (TSP) based on the Teamwork method. So-calledexperts and specialists work independently and in parallel for given time pe-riods. For TSP, specialists are tour construction algorithms and experts usemodified genetic algorithms in which after each application of a genetic operatorthe resulting tour is locally optimized before it is added to the population. Aftera given time period the work of each expert and specialist is judged by a referee.A new start population, including selected individuals from each expert and spe-cialist, is generated by the supervisor, based on the judgments of the referees.Our system is able to find better tours than each of the experts or specialistsworking alone. Also results comparable to those of single runs can be found muchfaster by a team.

Rules are an important knowledge representation formalism in constructive problem solving. On the other hand, object orientation is an essential key technology for maintaining large knowledge bases as well as software applications. Trying to take advantage of the benefits of both paradigms, we integrated Prolog and Smalltalk to build a common base architecture for problem solving. This approach has proven to be useful in the development of two knowledge-based systems for planning and configuration design (CAPlan and Idax). Both applications use Prolog as an efficient computational source for the evaluation of knowledge represented as rules.

Retrieval of cases is one important step within the case-based reasoning paradigm. We propose an improvement of this stage in the process model for finding most similar cases with an average effort of O[log2n], n number of cases. The basic idea of the algorithm is to use the heterogeneity of the search space for a density-based structuring and to employ this precomputed structure, a k-d tree, for efficient case retrieval according to a given similarity measure sim. In addition to illustrating the basic idea, we present the expe- rimental results of a comparison of four different k-d tree generating strategies as well as introduce the notion of virtual bounds as a new one that significantly reduces the retrieval effort from a more pragmatic perspective. The presented approach is fully implemented within the (Patdex) system, a case-based reasoning system for diagnostic applications in engineering domains.

We present the adaptation process in a CBR application for decision support in the domain of industrial supervision. Our approach uses explanations to approximate relations between a problem description and its solution, and the adaptation process is guided by these explanations (a more detailed presentation has been done in [4]).

The paper explores the role of artificial intelligence techniques in the development of an enhanced software project management tool, which takes account of the emerging requirement for support systems to address the increasing trend towards distributed multi-platform software development projects. In addressing these aims this research devised a novel architecture and framework for use as the basis of an intelligent assistance system for use by software project managers, in the planning and managing of a software project. This paper also describes the construction of a prototype system to implement this architecture and the results of a series of user trials on this prototype system.

Requirements engineering (RE) is a necessary part of the software development process, as it helps customers and designers identify necessary system requirements. If these stakeholders are separated by distance, we argue that a distributed groupware environment supporting a cooperative requirements engineering process must be supplied that allows them to negotiate software requirements. Such a groupware environment must support aspects of joint work relevant to requirements negotiation: synchronous and asynchronous collaboration, telepresence, and teledata. It should also add explicit support for a structured RE process, which includes the team's ability to discuss multiple perspectives during requirements acquisition and traceability. We chose the TeamWave software platform as an environment that supplied the basic collaboration capabilities, and tailored it to fit the specific needs of RE.

To prove difficult theorems in a mathematical field requires substantial know-ledge of that field. In this paper a frame-based knowledge representation formalismis presented, which supports a conceptual representation and to a large extent guar-antees the consistency of the built-up knowledge bases. We define a semantics ofthe representation by giving a translation into the underlaying logic.

We tested the GYROSTAR ENV-05S. This device is a sensor for angular velocity. There- fore the orientation must be calculated by integration of the angular velocity over time. The devices output is a voltage proportional to the angular velocity and relative to a reference. The test where done to find out under which conditions it is possible to use this device for estimation of orientation.

This paper addresses two modi of analogical reasoning. Thefirst modus is based on the explicit representation of the justificationfor the analogical inference. The second modus is based on the repre-sentation of typical instances by concept structures. The two kinds ofanalogical inferences rely on different forms of relevance knowledge thatcause non-monotonicity. While the uncertainty and non-monotonicity ofanalogical inferences is not questioned, a semantic characterization ofanalogical reasoning has not been given yet. We introduce a minimalmodel semantics for analogical inference with typical instances.

The paper focuses on the problem of trajectory planning of flexible redundant robot manipulators (FRM) in joint space. Compared to irredundant flexible manipulators, FRMs present additional possibilities in trajectory planning due to their kinematics redundancy. A trajectory planning method to minimize vibration of FRMs is presented based on Genetic Algorithms (GAs). Kinematics redundancy is integrated into the presented method as a planning variable. Quadrinomial and quintic polynomials are used to describe the segments which connect the initial, intermediate, and final points in joint space. The trajectory planning of FRMs is formulated as a problem of optimization with constraints. A planar FRM with three flexible links is used in simulation. A case study shows that the method is applicable.

We present first steps towards fully automated deduction that merely requiresthe user to submit proof problems and pick up results. Essentially, this necessi-tates the automation of the crucial step in the use of a deduction system, namelychoosing and configuring an appropriate search-guiding heuristic. Furthermore,we motivate why learning capabilities are pivotal for satisfactory performance.The infrastructure for automating both the selection of a heuristic and integra-tion of learning are provided in form of an environment embedding the "core"deduction system.We have conducted a case study in connection with a deduction system basedon condensed detachment. Our experiments with a fully automated deductionsystem 'AutoCoDe' have produced remarkable results. We substantiate Au-toCoDe's encouraging achievements with a comparison with the renowned the-orem prover Otter. AutoCoDe outperforms Otter even when assuming veryfavorable conditions for Otter.

In order to improve the quality of software systems and to set up a more effective process for their development, many attempts have been made in the field of software engineering. Reuse of existing knowledge is seen as a promising way to solve the outstanding problems in this field. In previous work we have integrated the design pattern concept with the formal design language SDL, resulting in a certain kind of pattern formalization. For the domain of communication systems we have also developed a pool of SDL patterns with an accompanying process model for pattern application. In this paper we present an extension that combines the SDL pattern approach with the experience base concept. This extension supports a systematic method for empirical evaluation and continuous improvement of the SDL pattern approach. Thereby the experience base serves as a repository necessary for effective reuse of the captured knowledge. A comprehensive usage scenario is described which shows the advantages of the combined approach. To demonstrate its feasibility, first results of a research case study are given.

Although several systematic analyses of existing approaches to adaptation have been published recently, a general formal adaptation framework is still missing. This paper presents a step into the direction of developing such a formal model of transformational adaptation. The model is based on the notion of the quality of a solution to a problem, while quality is meant in a more general sense and can also denote some kind of appropriateness, utility, or degree of correctness. Adaptation knowledge is then defined in terms of functions transforming one case into a successor case. The notion of quality provides us with a semantics for adaptation knowledge and allows us to define terms like soundness, correctness and completeness. In this view, adaptation (and even the whole CBR process) appears to be a special instance of an optimization problem.

This paper focuses on the issues involved when multiple mobile agents interact in multiagent systems. The application is an intelligent agent market place, where buyer and seller agents cooperate and compete to process sales transactions for their owners. The market place manager acts as afacilitator by giving necessary information to agents and managing communication between agents, and also as a mediator by proposing solutions to agents or stopping them to get into infinite loops bargaining back and forth.The buyer and seller agents range from using hardcoded logic to rule-based inferencing in their negotiation strategies. However these agents must support some communication skills using KQML or FIPA-ACL.So in contrast with other approaches to multiagent negotiation, we introduce an explicit mediator (market place manager) into the negotiation, and we propose a negotiation strategy based on dependence theory [1] implemented by our best buyers and best sellers.

In this paper we are interested in an algebraic specification language that (1) allowsfor sufficient expessiveness, (2) admits a well-defined semantics, and (3) allows for formalproofs. To that end we study clausal specifications over built-in algebras. To keep thingssimple, we consider built-in algebras only that are given as the initial model of a Hornclause specification. On top of this Horn clause specification new operators are (partially)defined by positive/negative conditional equations. In the first part of the paper wedefine three types of semantics for such a hierarchical specification: model-theoretic,operational, and rewrite-based semantics. We show that all these semantics coincide,provided some restrictions are met. We associate a distinguished algebra A spec to ahierachical specification spec. This algebra is initial in the class of all models of spec.In the second part of the paper we study how to prove a theorem (a clause) valid in thedistinguished algebra A spec . We first present an abstract framework for inductive theoremprovers. Then we instantiate this framework for proving inductive validity. Finally wegive some examples to show how concrete proofs are carried out.This report was supported by the Deutsche Forschungsgemeinschaft, SFB 314 (D4-Projekt)

This paper shows how a new approach to theorem provingby analogy is applicable to real maths problems. This approach worksat the level of proof-plans and employs reformulation that goes beyondsymbol mapping. The Heine-Borel theorem is a widely known result inmathematics. It is usually stated in R 1 and similar versions are also truein R 2 , in topology, and metric spaces. Its analogical transfer was proposedas a challenge example and could not be solved by previous approachesto theorem proving by analogy. We use a proof-plan of the Heine-Boreltheorem in R 1 as a guide in automatically producing a proof-plan of theHeine-Borel theorem in R 2 by analogy-driven proof-plan construction.

Most automated theorem provers suffer from the problem that theycan produce proofs only in formalisms difficult to understand even forexperienced mathematicians. Efforts have been made to transformsuch machine generated proofs into natural deduction (ND) proofs.Although the single steps are now easy to understand, the entire proofis usually at a low level of abstraction, containing too many tedioussteps. Therefore, it is not adequate as input to natural language gen-eration systems.To overcome these problems, we propose a new intermediate rep-resentation, called ND style proofs at the assertion level . After illus-trating the notion intuitively, we show that the assertion level stepscan be justified by domain-specific inference rules, and that these rulescan be represented compactly in a tree structure. Finally, we describea procedure which substantially shortens ND proofs by abstractingthem to the assertion level, and report our experience with furthertransformation into natural language.

We present an empirical study of mathematical proofs by diagonalization, the aim istheir mechanization based on proof planning techniques. We show that these proofs canbe constructed according to a strategy that (i) finds an indexing relation, (ii) constructsa diagonal element, and (iii) makes the implicit contradiction of the diagonal elementexplicit. Moreover we suggest how diagonal elements can be represented.

This paper presents a brief overview of the INRECA-II methodology for building and maintaining CBR applications. It is based on the experience factory and the software process modeling approach from software engineering. CBR development and maintenance experience is documented using software process models and stored in a three-layered experience packet.

In this report we present a case study of employing goal-oriented heuristics whenproving equational theorems with the (unfailing) Knut-Bendix completion proce-dure. The theorems are taken from the domain of lattice ordered groups. It will bedemonstrated that goal-oriented (heuristic) criteria for selecting the next critical paircan in many cases significantly reduce the search effort and hence increase per-formance of the proving system considerably. The heuristic, goalADoriented criteriaare on the one hand based on so-called "measures" measuring occurrences andnesting of function symbols, and on the other hand based on matching subterms.We also deal with the property of goal-oriented heuristics to be particularly helpfulin certain stages of a proof. This fact can be addressed by using them in a frame-work for distributed (equational) theorem proving, namely the "teamwork-method".

Orderings on polynomial interpretations of operators represent a powerful technique for proving thetermination of rewriting systems. One of the main problems of polynomial orderings concerns thechoice of the right interpretation for a given rewriting system. It is very difficult to develop techniquesfor solving this problem. Here, we present three new heuristic approaches: (i) guidelines for dealingwith special classes of rewriting systems, (ii) an algorithm for choosing appropriate special polynomialsas well as (iii) an extension of the original polynomial ordering which supports the generation ofsuitable interpretations. All these heuristics will be applied to examples in order to illustrate theirpractical relevance.

The notion of formal description techniques for timed systems (T-FDTs) has been introduced in [EDK98a] to provide a unifying framework for description techniques that are formal and that allow to describe the ongoing behavior of systems. In this paper we show that three well known temporal logics, MTL, MTL-R , and CTL*, can be embedded in this framework. Moreover, we provide evidence that a large number of dioeerent kinds of temporal logics can be considered as T-FDTs.

We transform a user-friendly formulation of aproblem to a machine-friendly one exploiting the variabilityof first-order logic to express facts. The usefulness of tacticsto improve the presentation is shown with several examples.In particular it is shown how tactical and resolution theoremproving can be combined.

Verfahren des Maschinellen Lernens haben heute eine Reife erreicht, die zu ersten erfolgreichen industriellen Anwendungen geführt hat. In der Prozessdiagnose und -steuerung ermöglichen Lernverfahren die Klassifikation und Bewertung von Betriebszuständen, d.h. eine Grobmodellierung eines Prozesses, wenn dieser nicht oder nur teilweise mathematisch beschreibbar ist. Ausserdem gestatten Lernverfahren die automatische Generierung von Klassifizierungsprozeduren, die deterministisch abgearbeitet werden und daher für die Belange der Echtzeitdiagnose und -steuerung u.U. zeiteffektiver als Inferenzmechanismen auf logischer bzw. Produktionsregelbasis sind, da letztere immer mit zeitaufwendigen Suchprozessen verbunden sind.

We study the combination of the following already known ideas for showing confluence ofunconditional or conditional term rewriting systems into practically more useful confluence criteria forconditional systems: Our syntactic separation into constructor and non-constructor symbols, Huet's intro-duction and Toyama's generalization of parallel closedness for non-noetherian unconditional systems, theuse of shallow confluence for proving confluence of noetherian and non-noetherian conditional systems, theidea that certain kinds of limited confluence can be assumed for checking the fulfilledness or infeasibilityof the conditions of conditional critical pairs, and the idea that (when termination is given) only primesuperpositions have to be considered and certain normalization restrictions can be applied for the sub-stitutions fulfilling the conditions of conditional critical pairs. Besides combining and improving alreadyknown methods, we present the following new ideas and results: We strengthen the criterion for overlayjoinable noetherian systems, and, by using the expressiveness of our syntactic separation into constructorand non-constructor symbols, we are able to present criteria for level confluence that are not criteria forshallow confluence actually and also able to weaken the severe requirement of normality (stiffened withleft-linearity) in the criteria for shallow confluence of noetherian and non-noetherian conditional systems tothe easily satisfied requirement of quasi-normality. Finally, the whole paper also gives a practically usefuloverview of the syntactic means for showing confluence of conditional term rewriting systems.

A growing share of all software development project work is being done by geographically distributed teams. To satisfy shorter product design cycles, expert team members for a development project may need to be r ecruited globally. Yet to avoid extensive travelling or r eplacement costs, distributed project work is preferred. Current-generation software engineering tools and ass ociated systems, processes, and methods were for the most part developed to be used within a single enterprise. Major innovations have lately been introduced to enable groupware applications on the Internet to support global collaboration. However, their deployment for distributed software projects requires further research. In partic ular, groupware methods must seamlessly be integrated with project and product management systems to make them attractive for industry. In this position paper we outline the major challenges concerning distributed (virtual) software projects. Based on our experiences with software process modeling and enactment environments, we then propose approaches to solve those challenges.

SPIN-NFDS Learning and Preset Knowledge for Surface Fusion - A Neural Fuzzy Decision System -
(1993)

The problem to be discussed in this paper may be characterized in short by the question: "Are these two surface fragments belonging together (i.e. belonging to the same surface)?" The presented techniques try to benefit from some predefined knowledge as well as from the possibility to refine and adapt this knowledge according to a (changing) real environment, resulting in a combination of fuzzy-decision systems and neural networks. The results are encouraging (fast convergence speed, high accuracy), and the model might be used for a wide range of applications. The general frame surrounding the work in this paper is the SPIN- project, where emphasis is on sub-symbolic abstractions, based on a 3-d scanned environment.

This paper refers to the problem of adaptability over an infinite period of time, regarding dynamic networks. A never ending flow of examples have to be clustered, based on a distance measure. The developed model is based on the self-organizing feature maps of Kohonen [6], [7] and some adaptations by Fritzke [3]. The problem of dynamic surface classification is embedded in the SPIN project, where sub-symbolic abstractions, based on a 3-d scanned environment is being done.

Viele Entwicklungsprozesse, wie sie z.B. beim Entwurf von grossen Softwaresystemen benötigt werden, basieren in erster Linie auf dem Wissen der mit der Entwicklung betrauten Mitarbeiter. Mit wachsender Komplexität der Entwurfsaufgaben und mit wachsender Anzahl der Mitarbeiter in einem Projekt wird die Koordination und Verteilung dieses Wissens immer problematischer. Aus diesem Grund versucht man zunehmend, das Wissen der Mitarbeiter in elektronischer Form, d.h. in Rechnern zu speichern und zu verwalten. Dadurch, dass der Entwurf eines komplexen Systems ebenfalls am Rechner modelliert wird, steht benötigtes Wissen sofort zur Verfügung und kann zur Entscheidungsunterstützung herangezogen werden. Gerade bei der Planung grosser Projekte stehen jedoch oft Entscheidungen aus, die erst später, während der Abwicklung getroffen werden können. Da gängige Workflow-Management-System zumeist eine komplette Modellierung verlangen, bevor die Abwicklung eines Projektmodells beginnen kann, habt sich dieser Ansatz gerade für umfangreiche Projekte als eher ungeeignet herausgestellt.

A non-trivial real-time requirement obeying a pattern that can be foundin various instantiations in the application domain building automation, and which is therefore called generic, is investigated in detail. Starting point is a description of a real-time problem in natural language augmented by a diagram, in a style often found in requirements documents. Step by step, this description is made more precise and finally transformed into a surprisingly concise formal specification, written in real-time temporal logic with customized operators. Wereason why this formal specification precisely captures the original description- as far as this is feasible due to the lack of precision of natural language.

Software Products As Objects
(1997)

This paper describes our experiences in modeling entire software products (trees of software files) as objects. Container pnodes (product nodes) have user-defined Internetunique names, data types, and methods (operations). Pnodes can contain arbitrary collections of software files that represent programs, libraries, documents, or other software products. Pnodes can contain multiple software products, so that header files, libraries, and program products may all be stored within one pnode. Pnodes can contain views that list other pnodes in order to form large conceptual structures of pnodes. Typical pnode -object methods include: fetching and storing into version controlled repositories; dynamic analysis of pnode contents to generate makefiles of arbitrary complexity; local automated build operations; Internet-scalable distributed repository synchroni- zations; Internet-scalable, multi-platform, distributed build operations; extraction and generation of online API documen- tation, spell checking of document pnodes, and so on. Since methods are user-defined, they can be arbitrarily complex. Modelling software products as objects provides a large amount of effort leverage, since one person can define the methods and many people can use them in extensively automated ways.

Da gerade in der heutigen Zeit viele zusammenarbeitende Softwareentwickler benötigt werden, um immer komplexer werdende Applikationen zu entwerfen, geht der Trend mehr und mehr in die Richtung des räumlich getrennten Arbeitens. Begünstigt wird diese Entwicklung nicht zuletzt durch die Möglichkeiten der Kommunikation und des Datenaustauschs, die durch das Internet geboten werden. Auf dieser Basis sollen Werkzeuge konzipiert und entwickelt werden, die eine effiziente verteilte Softwareentwicklung ermöglichen. Die Nutzung des Internet zu diesem Zweck löst das Verbindungsproblem für sehr große Entfernungen, die Nutzung von Webservern und -browsern wird der Anforderung der Betriebssystemunabhängigkeit und der Realisierung der Verteiltheit im Sinne des Client/Server-Prinzips gerecht. Unter dem Oberbegriff "Software Configuration Management" versteht man die Menge aller Aufgaben, die bei der Produktverwaltung im Bereich der Softwareherstellung anfallen. In dieser Ausarbeitung sollen zunächst die Anforderungen an ein webbasiertes SCM-System formuliert, einige technische Möglichkeiten genannt und verschiedene existierende SCM-Produkte, die eine Web-Schnittstelle bieten auf die Anforderungen überprüft und miteinander verglichen werden.

Techniques for modular software design are presented applying software agents. The conceptual designs are domain independent and make use of specificdomain aspects applying Multiagent AI. The stages of conceptualization, design and implementation are defined by new techniques coordinated by objects. Software systemsare designed by knowledge acquisition, specification, and multiagent implementations.

The development of software products has become a highly cooperative and distributed activity involving working groups at geographically distinct places. These groups show an increasing mobility and a very flexible organizational structure. Process methodology and technology have to take such evolutions into account. A possible direction for the emergence of new process technology and methodology is to take benefit from recent advances within multiagent systems engineering : innovative methodologies for adaptable and autonomous architectures; they exhibit interesting features to support distributed software processes.

SmallSync, an internet event synchronizer, is intended to provide a monitoring and visualization methodology for permitting simultaneous analysis and control of multiple remote processes on the web. The current SmallSync includes: (1) a mechanism to multicast web window-based commands, message passing events and process execution events among processes; (2) an event synchronizer to allow concurrent execution of some functions on multiple machines; (3) a means to report when these events cause errors in the processes; and (4) ad hoc visualization of process states using existing visualizers.

Patdex is an expert system which carries out case-based reasoning for the fault diagnosis of complex machines. It is integrated in the Moltke workbench for technical diagnosis, which was developed at the university of Kaiserslautern over the past years, Moltke contains other parts as well, in particular a model-based approach; in Patdex where essentially the heuristic features are located. The use of cases also plays an important role for knowledge acquisition. In this paper we describe Patdex from a principal point of view and embed its main concepts into a theoretical framework.

Object-oriented case representations require approaches for similarity assessment that allow to compare two differently structured objects, in particular, objects belonging to different object classes. Currently, such similarity measures are developed more or less in an ad-hoc fashion. It is mostly unclear, how the structure of an object-oriented case model, e.g., the class hierarchy, influences similarity assessment. Intuitively, it is obvious that the class hierarchy contains knowledge about the similarity of the objects. However, how this knowledge relates to the knowledge that could be represented in similarity measures is not obvious at all. This paper analyzes several situations in which class hierarchies are used in different ways for case modeling and proposes a systematic way of specifying similarity measures for comparing arbitrary objects from the hierarchy. The proposed similarity measures have a clear semantics and are computationally inexpensive to compute at run-time.

Simultaneous quantifier elimination in sequent calculus is an improvement over the well-known skolemization. It allows a lazy handling of instantiations as well as of the order of certain reductions. We prove the soundness of a sequent calculus which incorporates a rule for simultaneous quantifier elimination. The proof is performed by semantical arguments and provides some insights into the dependencies between various formulas in a sequent.

Self-localization in unknown environments respectively correlation of current and former impressions of the world is an essential ability for most mobile robots. The method,proposed in this article is the construction of a qualitative, topological world model as a basis for self-localization. As a central aspect the reliability regarding error-tolerance and stability will be emphasized. The proposed techniques demand very low constraints for the kind and quality of the employed sensors as well as for the kinematic precisionof the utilized mobile platform. Hard real-time constraints can be handled due to the low computational complexity. The principal discussions are supported by real-world experiments with the mobile robot.

This article will discuss a qualitative, topological and robust world-modelling technique with special regard to navigation-tasks for mobile robots operating in unknownenvironments. As a central aspect, the reliability regarding error-tolerance and stability will be emphasized. Benefits and problems involved in exploration, as well as in navigation tasks, are discussed. The proposed method demands very low constraints for the kind and quality of the employed sensors as well as for the kinematic precision of the utilized mobile platform. Hard real-time constraints can be handled due to the low computational complexity. The principal discussions are supported by real-world experiments with the mobile robot

Die Domäne der Operationsroboter liegt heute in Fräsarbeiten an knöchernen Strukturen. Da Roboter über eine extreme Präzision verfügen und nicht ermüden bietet sich ihr Einsatz ins-besondere bei langwierigen und zugleich hochpräzisen Fräsvorgängen im Bereich der later-alen Schädelbasis an. Aus diesem Grunde wurde ein Verfahren entwickelt, welches aus einer geometrischen Beschreibung des Implantates eine geeignete Fräsbahn errechnet und eine kraftgeregelte Prozesskontrolle des Fräsvorganges implementiert. Mit einem 6*achsigen Knickarmroboter erfolgten die Untersuchungen primär an Tierpräparaten und zur Optimierung an Felsenbeinpräparaten.

Besides the work in the field of manipulating rigid objects, currently, there are several research and development activities going on in the field of manipulating non-rigid or deformable objects. Several papers have been published on international conferences in this field from various projects and countries. But there has been no comprehensive work which provides both a representative overview of the state of the art and identifies the important aspects in this field. Thus, we collected these activities and invited the corresponding working groups to present an overview of their research. Altogether, nineteen authors coming from Japan, Germany, Italy, Greece, United Kingdom, and Australia contributed to this book. Their research work covers all the different aspects that occur when manipulating deformable objects. The contributions can be characterized and grouped by the following four aspects: * object modeling and simulation, * planning and control strategies, * collaborative systems, and * applications and industrial experiences. In the following, we give a short motivation and overview of the single chapters of the book. The simulation of deformable objects is one way to approach the problem of manipulating these objects by robots. Based on a physical model of the object and the occurring constraints, the resulting object shape is calculated. In Chapter 2, Hirai presents an energy-based approach, where the internal energy under the geometric constraints is minimized. Frugoli et al. introduce a force-based approach, where the forces between discrete particles are minimized meeting given constraints. Finally, Remde and Henrich extend the energy-based approach to plastic deformation and give a solution of the inverse simulation problem. Even if the object behavior is predicted by simulation, there is still the question of how to control the robot during a single manipulation operation. An additional question is how to retrieve an overall plan for the concatenated manipulation operations. In Chapter 3, Wada investigates the control problems when positioning multiple points of a planar deformable object. McCarrager proposes a control scheme exploiting the flexibility, rather than minimizing it. Abegg et al. use a simple contact state model to describe typical assembly tasks and to derive robust manipulation primitives. Finally, Ono presents an automatic sewing system and suggests a strategy for unfolding fabric. In several manipulation tasks, it is reasonable to apply more than one robot. Especially in cases, where the deformable object has to take a specific shape. Since the robots working at the same object are influencing each other, different control algorithms have to be introduced. In Chapter 4, Yoshida and Kosuge investigates this problem for the task of bending a sheet of metal and exploits the relation ship between the static object deformation and the bending moments. Tanner and Kyriakopoulos regard the deformable object as underactuated mechanical system and make use of the existence of non-holonomic constraints. Both approaches model the deformable object as finite elements. All of the above aspects have their counterpart in different applications and industrial experiences. In Chapter 5, Rizzi et al. present test cases and applications of their approach to simulate the manipulation of fabric, wires, cables, and soft bags. Buckingham and Graham give an overview of two European projects processing white fish including locating, gripping, and deheading the fish. Maruyama outlines the three development phases of a robot system for performing outage-free maintenance of live-line power supply in Japan. Finally, Kämper presents the development of a flexible automatic cabling unit for the wiring of long-tube lighting with plug components.

Reusing Proofs
(1999)

We develop a learning component for a theorem prover designed for verifying statements by mathematical induction. If the prover has found a proof, it is analyzed yielding a so-called catch. The catch provides the features of the proof which are relevant for reusing it in subsequent verification tasks and may also suggest useful lemmata. Proof analysis techniques for computing the catch are presented. A catch is generalized in a certain sense for increasing the reusability of proofs. We discuss problems arising when learning from proofs and illustrate our method by several examples.

This paper describes a declarative approach forencoding the plan operators in proof planning,the so-called methods. The notion of methodevolves from the much studied concept of a tac-tic and was first used by A. Bundy. Signific-ant deductive power has been achieved withthe planning approach towards automated de-duction; however, the procedural character ofthe tactic part of methods hinders mechanicalmodification. Although the strength of a proofplanning system largely depends on powerfulgeneral procedures which solve a large class ofproblems, mechanical or even automated modi-fication of methods is necessary, since methodsdesigned for a specific type of problems willnever be general enough. After introducing thegeneral framework, we exemplify the mechan-ical modification of methods via a particularmeta-method which modifies methods by trans-forming connectives to quantifiers.

We examine an approach for demand-driven cooperative theorem proving.We briefly point out the problems arising from the use of common success-driven cooperation methods, and we propose the application of our approachof requirement-based cooperative theorem proving. This approach allows for abetter orientation on current needs of provers in comparison with conventional co-operation concepts. We introduce an abstract framework for requirement-basedcooperation and describe two instantiations of it: Requirement-based exchangeof facts and sub-problem division and transfer via requests. Finally, we reporton experimental studies conducted in the areas superposition and unfailing com-pletion.The author was supported by the Deutsche Forschungsgemeinschaft (DFG).

Der Wissenserwerb erschwert bisher häufig den Einsatz wissensbasierter Systeme der Arbeitsplanerstellung in der industriellen Praxis. Die meisten Anwendungen gestatten nur das Erfassen und Editieren des durch aufwendige Erhebung, Systematisierung und Formulierung gewonnenen fachspezifischen Planungswissens. Im Rahmen eines DFG-Projektes soll die Anwendbarkeit bekannter maschineller Lernverfahren auf technologische Reihenfolge- und Zuordnungsprobleme im Rahmen der generierenden Arbeitsplanerstellung von Teilefertigungsprozessen im Maschinenbau nachgewiesen werden. Dazu wird ein Prototyp mit Hilfe eines verfügbaren Softwarewerkzeuges entwickelt, der das maschinelle Lernen aus vorgegebenen Beispielen ermöglichen und mit einem existierenden Prototypen der wissensbasierten Arbeistplanung kommunizieren soll. Der folgende Beitrag gibt einen Überblick über das mit Lernverfahren zu behandelnde Planungswissen und stellt mögliche Repräsentationsmöglichkeiten des Wissens zur Diskussion.

Emerging technologies such as the Internet, the World Wide Web, JavaTM technology, and software components, are changing the software business. Activities that have in the past been constrained by the need for intense information management increasingly involve cooperating organizations. Information management tools and techniques do not scale well in the face of this organizational complexity. An informal approach to information sharing, based largely on manual copying of information, cannot meet the demands of the task as size and complexity increase. Formal approaches to sharing information are based on groupware tools, but cooperating organizations do not always enjoy the trust or commonality of sophisticated infrastructure, methods, and skills that this approach requires. Bridging the gap requires a simple, loosely coupled, highly flexible strategy for information sharing. Extensive information relevant to different parts of the software life cycle should be interconnected in a simple, easily described way; such connections should permit selective information sharing by a variety of tools and in a variety of collaboration modes that vary in the amount of organizational coupling they require.

We investigate restricted termination and confluence properties of term rewritADing systems, in particular weak termination and innermost termination, and theirinterrelation. New criteria are provided which are sufficient for the equivalenceof innermost / weak termination and uniform termination of term rewriting sysADtems. These criteria provide interesting possibilities to infer completeness, i.e.termination plus confluence, from restricted termination and confluence properADties.Using these basic results we are also able to prove some new results aboutmodular termination of rewriting. In particular, we show that termination ismodular for some classes of innermost terminating and locally confluent termrewriting systems, namely for nonADoverlapping and even for overlay systems. Asan easy consequence this latter result also entails a simplified proof of the factthat completeness is a decomposable property of soADcalled constructor systems.Furthermore we show how to obtain similar results for even more general cases of(nonADdisjoint) combined systems with shared constructors and of certain hierarADchical combinations of systems with constructors. Interestingly, these modularityresults are obtained by means of a proof technique which itself constitutes a modADular approach.

Es wird das Lernen uniform rekursiv aufzählbarer Sprachfamilien anhand guter Beispiele untersucht und Unterschiede und Gemeinsamkeiten zum Lernen von rekursiven Sprachfamilien und rekursiven Funktionen aufgezeigt. Dem verwendeten Modell liegt das Lernen von Schülern mit einem Lehrer zugrunde. Es werden verschiedene Varianten vorgestellt, verglichen und teilweise auch charakterisiert, und versucht, mit Beispielen und anderen typischen Eigenschaften ein Gefühl für die Leistungsfähigkeit zu vermitteln. Unter anderem wird gezeigt, dass es nicht immer "universelle" gute Beispiele gibt, mit denen eine Sprachklasse in allen Situationen erklärt werden kann.

Like other industries, the aircraft industry is under high pressure to meet drastically increased customer goals for market price and flexibility. This while at the same time share holders request for short term profit guarantees. Daimler-Benz Aerospace Airbus has met this challenge using business process reengineering methods which led to total company restructuring from functional orientation to customer and product orientation. This paper will show how business process modelling techniques have been applied. Especially concurrent engineering methods are used to integrate the various disciplines involved from market analysts over design, commercial to industrialization staff.

The problem to be discussed here, is the usage of neural network clustering techniques on a mobile robot, in order to build qualitative topologic environment maps. This has to be done in realtime, i.e. the internal world model has to be adapted by the flow of sensor- samples without the possibility to stop this data-flow.Our experiments are done in a simulation environment as well as on a robot, called ALICE.

This paper discusses the benefits and drawbacks of caching and replication strategies in the WWW with respect to the Internet infrastructure. Bandwidth consumption, latency, and overall error rates are considered to be most important from a network point of view. The dependencies of these values with input parameters like degree of replication, document popularity, actual cache hit rates, and error rates are highlighted. In order to determine the influence of different caching and replication strategies on the behavior of a single proxy server with respect to these values, trace-based simulations are used. Since the overall effects of such strate- gies can hardly be decided with this approach alone, a mathematical model has been developed to deal with their influence on the network as a whole. Together, this two-tiered approach permits us to propose quantita- tive assessments on the influence different caching and replication proposals (are going to) have on the Inter- net infrastructure.

A lot of the human ability to prove hard mathematical theorems can be ascribedto a problem-specific problem solving know-how. Such knowledge is intrinsicallyincomplete. In order to prove related problems human mathematicians, however,can go beyond the acquired knowledge by adapting their know-how to new relatedproblems. These two aspects, having rich experience and extending it by need, can besimulated in a proof planning framework: the problem-specific reasoning knowledge isrepresented in form of declarative planning operators, called methods; since these aredeclarative, they can be mechanically adapted to new situations by so-called meta-methods. In this contribution we apply this framework to two prominent proofs intheorem proving, first, we present methods for proving the ground completeness ofbinary resolution, which essentially correspond to key lemmata, and then, we showhow these methods can be reused for the proof of the ground completeness of lockresolution.

This paper addresses analogy-driven auto-mated theorem proving that employs a sourceproof-plan to guide the search for a proof-planof the target problem. The approach presen-ted uses reformulations that go beyond symbolmappings and that incorporate frequently usedre-representations and abstractions. Severalrealistic math examples were successfully pro-cessed by our analogy-driven proof-plan con-struction. One challenge example, a Heine-Borel theorem, is discussed here. For this ex-ample the reformulaitons are shown step bystep and the modifying actions are demon-strated.

This paper outlines an implemented system called PROVERB that explains machine -found natural deduction proofs in natural language. Different from earlier works, we pursue a reconstructive approach. Based on the observation that natural deduction proofs are at a too low level of abstraction compared with proofs found in mathematical textbooks, we define first the concept of so-called assertion level inference rules. Derivations justified by these rules can intuitively be understood as the application of a definition or a theorem. Then an algorithm is introduced that abstracts machine-found ND proofs using the assertion level inference rules. Abstracted proofs are then verbalized into natural language by a presentation module. The most significant feature of the presentation module is that it combines standard hierarchical text planning and techniques that locally organize argumentative texts based on the derivation relation under the guidance of a focus mechanism. The behavior of the system is demonstrated with the help of a concrete example throughout the paper.

This paper outlines the linguistic part of an implemented system namedPROVERB[3] that transforms, abstracts, and verbalizes machine-found proofs innatural language. It aims to illustrate, that state-of-the-art techniques of natural language processing are necessary to produce coherent texts that resemble those found in typical mathematical textbooks, in contrast to the belief that mathematical texts are only schematic and mechanical.The verbalization module consists of a content planner, a sentence planner, and a syntactic generator. Intuitively speaking, the content planner first decides the order in which proof steps should be conveyed. It also some messages to highlight global proof structures. Subsequently, thesentence planner combines and rearranges linguistic resources associated with messages produced by the content planner in order to produce connected text. The syntactic generat or finally produces the surface text.

This paper describes the linguistic part of a system called PROVERB, which transforms, abstracts,and verbalizes machine-found proofs into formatedtexts. Linguistically, the architecture of PROVERB follows most application oriented systems, and is a pipelined control of three components. Its macroplanner linearizes a proof and plans mediating communicative acts by employing a combination of hierarchical planning and focus-guided navigation. The microplanner then maps communicative acts and domain concepts into linguistic resources, paraphrases and aggregates such resources to producethe final Text Structure. A Text Structure contains all necessary syntactic information, and can be executed by our realizer into grammatical sentences. The system works fully automatically and performs particularly well for textbook size examples.

The paper addresses two problems of comprehensible proof presentation, the hierarchically structured presentation at the level of proof methods and different presentation styles of construction proofs. It provides solutions for these problems that can make use of proof plans generated by an automated proof planner.

We first show that ground term-rewriting systems can be completed in apolynomial number of rewriting steps, if the appropriate data structure for termsis used. We then apply this result to study the lengths of critical pair proofs innon-ground systems, and obtain bounds on the lengths of critical pair proofsin the non-ground case. We show how these bounds depend on the types ofinference steps that are allowed in the proofs.

Proof planning is an alternative methodology to classical automated theorem prov-ing based on exhausitve search that was first introduced by Bundy [8]. The goal ofthis paper is to extend the current realm of proof planning to cope with genuinelymathematical problems such as the well-known limit theorems first investigated for au-tomated theorem proving by Bledsoe. The report presents a general methodology andcontains ideas that are new for proof planning and theorem proving, most importantlyideas for search control and for the integration of domain knowledge into a general proofplanning framework. We extend proof planning by employing explicit control-rules andsupermethods. We combine proof planning with constraint solving. Experiments showthe influence of these mechanisms on the performance of a proof planner. For instance,the proofs of LIM+ and LIM* have been automatically proof planned in the extendedproof planner OMEGA.In a general proof planning framework we rationally reconstruct the proofs of limittheorems for real numbers (IR) that were first computed by the special-purpose programreported in [6]. Compared with this program, the rational reconstruction has severaladvantages: It relies on a general-purpose problem solver; it provides high-level, hi-erarchical representations of proofs that can be expanded to checkable ND-proofs; itemploys declarative contol knowledge that is modularly organized.

This paper outlines an implemented system named PROVERBthat transforms and abstracts machine-found proofs to natural deduction style proofs at an adequate level of abstraction and then verbalizesthem in natural language. The abstracted proofs, originally employedonly as an intermediate representation, also prove to be useful for proofplanning and proving by analogy.

Point-to-Point Trajectory Planning of Flexible Redundant Robot Manipulators Using Genetic Algorithms
(2001)

The paper focuses on the problem of point-to-point trajectory planning for flexible redundant robot manipulators (FRM) in joint space. Compared with irredundant flexible manipulators, a FRM possesses additional possibilities during point-to-point trajectory planning due to its kinematics redundancy. A trajectory planning method to minimize vibration and/or executing time of a point-to-point motion is presented for FRM based on Genetic Algorithms (GAs). Kinematics redundancy is integrated into the presented method as planning variables. Quadrinomial and quintic polynomial are used to describe the segments that connect the initial, intermediate, and final points in joint space. The trajectory planning of FRM is formulated as a problem of optimization with constraints. A planar FRM with three flexible links is used in simulation. Case studies show that the method is applicable.

This article presents contributions in the field of path planning for industrial robots with 6 degrees of freedom. This work presents the results of our research in the last 4 years at the Institute for Process Control and Robotics at the University of Karlsruhe. The path planning approach we present works in an implicit and discretized C-space. Collisions are detected in the Cartesian workspace by a hierarchical distance computation. The method is based on the A* search algorithm and needs no essential off-line computation. A new optimal discretization method leads to smaller search spaces, thus speeding up the planning. For a further acceleration, the search was parallelized. With a static load distribution good speedups can be achieved. By extending the algorithm to a bidirectional search, the planner is able to automatically select the easier search direction. The new dynamic switching of start and goal leads finally to the multi-goal path planning, which is able to compute a collision-free path between a set of goal poses (e.g., spot welding points) while minimizing the total path length.

This paper provides a description of PLATIN. With PLATIN we present an imple-mented system for planning inductive theorem proofs in equational theories that arebased on rewrite methods. We provide a survey of the underlying architecture ofPLATIN and then concentrate on details and experiences of the current implementa-tion.

Planverfahren
(1999)

Planning for manufacturing workpieces is a complex task that requires the interaction of a domain-specific reasoner and a generic planning mechanism. In this paper we present an architecture for organizing the case base that is based on the information provided by a generic problem solver. A retrieval procedure is then presented that uses the information provided by the domain-specific reasoner in order to improve the accuracy of the cases retrieved. However, it is not realistic to suppose that the case retrieved will entirely fit into the new problem. We present a replay procedure to obtain a partial solution that replays not only the valid decisions taken for solving the case, but also justifications of rejected decisions made during the problem solving process. As a result, those completion alternatives of the partial solution are discarded that are already known to be invalid from the case.

We describe a hybrid architecture supporting planning for machining workpieces. The archi- tecture is built around CAPlan, a partial-order nonlinear planner that represents the plan already generated and allows external control decision made by special purpose programs or by the user. To make planning more efficient, the domain is hierarchically modelled. Based on this hierarchical representation, a case-based control component has been realized that allows incremental acquisition of control knowledge by storing solved problems and reusing them in similar situations.

We describe a hybrid architecture supporting planning for machining workpieces. The architecture is built around CAPlan, a partial-order nonlinear planner that represents the plan already generated and allows external control decision made by special purpose programs or by the user. To make planning more efficient, the domain is hierarchically modelled. Based on this hierarchical representation, a case-based control component has been realized that allows incremental acquisition of control knowledge by storing solved problems and reusing them in similar situations.

This paper presents a new way to use planning in automated theorem provingby means of distribution. To overcome the problem that often subtasks fora proof problem can not be detected a priori (which prevents the use of theknown planning and distribution techniques) we use a team of experts that workindependently with different heuristics on the problem. After a certain amount oftime referees judge their results using the impact of the results on the behaviourof the expert and a supervisor combines the selected results to a new startingpoint.This supervisor also selects the experts that can work on the problem inthe next round. This selection is a reactive planning task. We outline whichinformation the supervisor can use to fulfill this task and how this informationis processed to result in a plan or to revise a plan. We also show that the useof planning for the assignment of experts to the team allows the system to solvemany different examples in an acceptable time with the same start configurationand without any consultation of the user.Plans are always subject to changeShin'a'in proverb

This report is a first attempt of formalizing the diagonalization proof technique.We give a strategy how to systematically construct diagonalization proofs: (i) findingan indexing relation, (ii) constructing a diagonal element, and (iii) making the implicitcontradiction of the diagonal element explicit. We suggest a declarative representationof the strategy and describe how it can be realized in a proof planning environment.

Abstraction is one of the most promising approaches to improve the performance of problem solvers. Abstraction by dropping sentences of a domain description - as used in most hierarchical planners - is known to be very representation dependent. To overcome these drawbacks, we propose a more general view of abstraction involving the change of representation language. We have developed a new abstraction methodology and a related sound and complete learning algorithm that allows the complete change of representation language of planning cases from concrete to abstract.

Due to the large variety of modern applications and evolving network technologies, a small number of general-purpose protocol stacks will no longer be sufficient. Rather, customization of communication protocols will play a major role. In this paper, we present an approach that has the potential to substantially reduce the effort for designing customized protocols. Our approach is based on the concept of design patterns, which is well-established in object oriented software development. We specialize this concept to communication protocols, and - in addition - use formal description techniques (FDTs) to specify protocol design patterns as well as rules for their instantiation and composition. The FDTs of our choice are SDL-92 and MSCs, which offer suitable language support. We propose an SDL pattern description template and relate pattern-based configuring of communication protocols to existing SDL methodologies. Particular SDL patterns and the configuring of a customized resource reservation protocol are presented in detail.

Im Bereich der Expertensysteme ist das Problemlösen auf der Basis von bekannten Fallbeispielen ein derzeit sehr aktuelles Thema. Auch für Diagnoseaufgaben gewinnt der fallbasierte Ansatz immer mehr an Bedeutung. In diesem Papier soll der im Rahmen des Moltke -Projektes1 an der Universität Kaiserslautern entwickelte fallbasierte Problemlöser Patdex/22 vorgestellt werden. Ein erster Prototyp, Patdex/1, wurde bereits 1988 entwickelt.

Paris (Plan Abstraction and Refinement in an Integrated System) [4, 2] is a domain independent case-based planning system which allows the flexible reuse of planning cases by abstraction and refinement. This approach is mainly inspired by the observation that reuse of plans must not be restricted to a single description level. In domains with a high variation in the problems, the reuse of past solutions must be achieved at various levels of abstraction.

We argue in this paper that sophisticated mi-croplanning techniques are required even formathematical proofs, in contrast to the beliefthat mathematical texts are only schematicand mechanical. We demonstrate why para-phrasing and aggregation significantly en-hance the flexibility and the coherence ofthe text produced. To this end, we adoptedthe Text Structure of Meteer as our basicrepresentation. The type checking mecha-nism of Text Structure allows us to achieveparaphrasing by building comparable combi-nations of linguistic resources. Specified interms of concepts in an uniform ontologicalstructure called the Upper Model, our se-mantic aggregation rules are more compactthan similar rules reported in the literature.

Distributed systems are an alternative to shared-memorymultiprocessors for the execution of parallel applications.PANDA is a runtime system which provides architecturalsupport for efficient parallel and distributed program-ming. PANDA supplies means for fast user-level threads,and for a transparent and coordinated sharing of objectsacross a homogeneous network. The paper motivates themajor architectural choices that guided our design. Theproblem of sharing data in a distributed environment isdiscussed, and the performance of appropriate mecha-nisms provided by the PANDA prototype implementation isassessed.

PANDA is a run-time package based on a very small operating system kernel which supports distributed applications written in C++. It provides powerful abstractions such as very efficient user-level threads, a uniform global address space, object and thread mobility, garbage collection, and persistent objects. The paper discusses the design ration- ales underlying the PANDA system. The fundamental features of PANDA are surveyed, and their implementation in the current prototype environment is outlined.

We present a novel approach to classification, based on a tight coupling of instancebased learning and a genetic algorithm. In contrast to the usual instance-based learning setting, we do not rely on (parts of) the given training set as the basis of a nearestneighbor classifier, but we try to employ artificially generated instances as concept prototypes. The extremely hard problem of finding an appropriate set of concept prototypes is tackled by a genetic search procedure with the classification accuracy on the given training set as evaluation criterion for the genetic fitness measure. Experiments with artificial datasets show that - due to the ability to find concise and accurate concept descriptions that contain few, but typical instances - this classification approach is considerably robust against noise, untypical training instances and irrelevant attributes. These favorable (theoretical) properties are corroborated using a number of hard real-world classification problems.

Recent studies on planning, comparing plan re-use and plan generation, have shown that both the above tasks may have the same degree of computational complexity, even if we deal with very similar problems. The aim of this paper is to show that the same kind of results apply also for diagnosis. We propose a theoretical complexity analysis coupled with some experimental tests, intended to evaluate the adequacy of adaptation strategies which re-use the solutions of past diagnostic problems in order to build a solution to the problem to be solved. Results of such analysis show that, even if diagnosis re-use falls into the same complexity class of diagnosis generation (they are both NP-complete problems), practical advantages can be obtained by exploiting a hybrid architecture combining case-based and modelbased diagnostic problem solving in a unifying framework.

For defining attribute types to be used in the case representation, taxonomies occur quite often. The symbolic values at any node of the taxonomy tree are used as attribute values in a case or a query. A taxonomy type represents a relationship between the symbols through their position within the taxonomy-tree which expresses knowledge about the similarity between the symbols. This paper analyzes several situations in which taxonomies are used in different ways and proposes a systematic way of specifying local similarity measures for taxonomy types. The proposed similarity measures have a clear semantics and are easy to compute at runtime.

This paper addresses the role of abstraction in case-based reasoning. We develop a general framework for reusing cases at several levels of abstraction, which is particularly suited for describing and analyzing existing and designing new approaches of this kind. We show that in synthetic tasks (e.g. configuration, design, and planning), abstraction can be successfully used to improve the efficiency of similarity assessment, retrieval, and adaptation. Furthermore, a case-based planning system, called Paris, is described and analyzed in detail using this framework. An empirical study done with Paris demonstrates significant advantages concerning retrieval and adaptation efficiency as well as flexibility of adaptation. Finally, we show how other approaches from the literature can be classified according to the developed framework.

To prove difficult theorems in a mathematical field requires substantial know-ledge of that field. In this thesis a frame-based knowledge representation formal-ism including higher-order sorted logic is presented, which supports a conceptualrepresentation and to a large extent guarantees the consistency of the built-upknowledge bases. In order to operationalize this knowledge, for instance, in anautomated theorem proving system, a class of sound morphisms from higher-orderinto first-order logic is given, in addition a sound and complete translation ispresented. The translations are bijective and hence compatible with a later proofpresentation.In order to prove certain theorems the comprehension axioms are necessary,(but difficult to handle in an automated system); such theorems are called trulyhigher-order. Many apparently higher-order theorems (i.e. theorems that arestated in higher-order syntax) however are essentially first-order in the sense thatthey can be proved without the comprehension axioms: for proving these theoremsthe translation technique as presented in this thesis is well-suited.

Freivalds, Karpinski and Smith [8] explored a special type of learning in the limit: identification of an unknown concept (function) by eliminating (erasing) all but one possible hypothesis (this type of learning is called co-learning). The motivation behind learning by erasing lies in the process of human and automated computer learning: often we can discard incorrect solutions much easier than to come up with the correct one. In Gödel numberings any learnable family can be learned by an erasing strategy. In this paper we concentrate on co-learning minimal programs. We show that co-learning of minimal programs, as originally defined is significantly weaker than learning minimal programs in Gödel numberings. In order to enhance the learning power