Robert Trappl

Personalities for Synthetic Actors: Current Issues and Some Perspectives

The content of the present volume depicts a lively area of research where a number of impressive results have already been achieved and successfully carried across to commercial applications. As illustrated in the overview given in the introductory chapter, these efforts were originally motivated by a wide variety of different objectives, which in turn fostered the exploration and adoption of a whole spectrum of personality engendering architectures and processes. Resuming and furthering this discussion, we will now proceed to analyse the current l imitations and shortcomings of the various approaches and point out possible directions to extend the cover functionalities or amend existing problems. Adding to that , we will give a brief description of some further lines of research which we expect to become of practical relevance in this area, thereby suggesting starting points for future directions of research, in which interdisciplinary efforts will continue to play an essential role. If we now pick up and continue to follow the thread laid down in the first chapter, we find two dimensions along which to structure our comments. The first one concerns the different kinds of basic approaches to how to create synthetic actors with personalities: here we will contrast lines of research applying detached drama-theoret ic techniques from the ou t s ide -where actors are seen as remotely controlled puppets engaging in fictitious in teract ions-to such that pursue the modeling of characters endowed with a self and which typically draw upon results from physics, biology, and the cognitive sciences. Within this second group, we will then make a further distinction: surface models are aimed at achieving external behaviours and display of traits which convey the impression of believable personalities, thus placing the emphasis on achieving the effect of a believable performance, if possible under preservation of real-time responsiveness and performance. On the other hand we find the deep generative cognitive models, in which issues such as emotions and personality are not considered and covered as ends to themselves, but rather are integrated in the respective architectures as necessary components for the achievement of rat ionali ty and social competence.

Knowledge structure definition for an expert system in primary medical care

This paper presents an outline of the program KSDS (Knowledge Structure Definition System) designed for the definition of the structure of a medical knowledge base. KSDS defines type and structure of the elements of a conceptual net. KSDS generates the internal structure and format of the knowledge base. Nodes and relations of this net are structured in a uniform fashion, consisting of four parts: identification, description of attributes, relations to concepts, and explanations. For each attribute there exists a list of primitives which restricts its type and possible values. The relation part is constructed referencing already defined relations of the structure base. If the structure is changed, KSDS takes care of the re-formating of the knowledge base. The system is currently being used for the definition and maintalnance of the knowledge base of an expert system in primary medical care.

MACHINE LEARNING AND CASE-BASED REASONING: THEIR POTENTIAL ROLE IN PREVENTING THE OUTBREAK OF WARS OR IN ENDING THEM

In a current project we investigate the potential contribution of Artificial Intelligence for the avoidance and termination of crises and wars. This paper reports some results obtained by analyzing international conflict databases using machine learning and case-based reasoning techniques.

VIE-LANG: A GERMAN LANGUAGE DIALOGUE SYSTEM

Abstract The paper gives a general description of VIE-LANG, a natural language understanding system for German. The components of the system are strongly based on semantics. The central component is a Structured Inheritance Network, which is the source of real world knowledge for all processes of the system. VIE-LANG is designed to engage in dialogues in German. Due to our focus of interest in methodology of representation we chose an approach independent of the intended domain. The designated area of application will be medicine.

A second order model of the optic generator potential and its relation to Stevens' power function

SummaryThe component PIII of the electroretinogram representing the optic generator potential was recorded after stimulation with short stimuli with different light intensity. It is shown that the impulse response function of a linear second order model with intensity-dependent coefficients can be well fit to the recordings. Two of its parameters, after logarithmic transformations, are linearly dependent on the luminance while the third parameter varies only within a small interval. It is therefore possible to describe the relation between PIII and luminance in a linearized second order model. Furthermore, both the type of the function relating the gain of the model to the luminance and its exponent are nearly identical with the psychophysic function relating luminance to subjective brightness. Further physiological implications are also considered.

A mathematical model of slow PIII of the ERG: Variation of stimulus duration

Light stimulation of a vertebrate retina induces an interruption of the Na’ inward current into the outer segments of photoreceptors (Penn and Hagins, 1969). Concomitantly [K’] decreases in the extracellular space of the photoreceptor layer (Tomita, 1976; Oakley and Green, 1976). It is hypothesized that the Na’ influx into photoreceptors is accompanied by a K‘ efBux. keeping a constant extracellular [K’] in the dark. Light reduces this outflow while the Na’,/K’ pump situated in the inner ‘segment membrane (Matsuura et al., 1978) continues to remove K’ from the extracellular space. It has been shown recently, that blocking the pump actually diminishes the light induced decrease of extracellular [K+] (Oakley et al., 1979). The [K’] sink seems to be mainly produced by rod activity and has been assumed to give rise to the Miiller cell hyperpolarization (Witkovsky et al., 1975) which can be recorded as slow P III in the gross ERG. Taking into account the different underlying mechanisms in the present investigation slow P III as recorded from an isolated rabbit eye bulb was modelled by a set of two exponential functions, which were fit independently to the potential changes occurring at light onset and offset respectively.

Agents across Cultures

The paper presents a short overview of the problems involved and solutions offered on the issue of adapting agents to different cultures. First, three different domains of human-agent interaction are distinguished – agent culture, user culture, and hybrid culture – in order to determine the degree to which IVAs have to adapt to cultural diversity. Turning to user culture in what follows, we outline the levels of agent architecture where such adaptation should be taken care of in the design process. We conclude with a couple of ”Agent Culture Hypotheses”.

Multi-agent systems and applications: 9th ECCAI advanced course, ACAI 2001 and agent link's 3rd European Agent Systems Summer School, EASSS 2001, Prague, Czech Republic, July 2-13, 2001 : selected tutorial papers

Foundations of Multi-agent Systems.- Perspectives on Organizations in Multi-agent Systems.- Multi-agent Infrastructure, Agent Discovery, Middle Agents for Web Services and Interoperation.- Logical Foundations of Agent-Based Computing.- Standardizing Agent Communication.- Standardizing Agent Interoperability: The FIPA Approach.- Distributed Problem Solving and Planning.- Automated Negotiation and Decision Making in Multiagent Environments.- Agents? Advanced Features for Negotiation and Coordination.- Social Behaviour, Meta-reasoning, and Learning.- Towards Heterogeneous Agent Teams.- Social Knowledge in Multi-agent Systems.- Machine Learning and Inductive Logic Programming for Multi-agent Systems.- Relational Reinforcement Learning.- From Statistics to Emergence: Exercises in Systems Modularity.- Emotions and Agents.- Applications.- Multi-agent Coordination and Control Using Stigmergy Applied to Manufacturing Control.- Virtual Enterprise Modeling and Support Infrastructures: Applying Multi-agent System Approaches.- Specialised Agent Applications.- Agent-Based Modelling of Ecosystems for Sustainable Resource Management.- Cooperating Physical Robots: A Lesson in Playing Robotic Soccer.- A Multi-agent Study of Interethnic Cooperation.

Facets of Emotions in Humans and Artifacts

During the last years it has become more and more clear that “the rational” and “the emotional” do not exclude each other, on the contrary, it is e.g. difficult for a person to make rational decisions if her or his brain structures responsible for experiencing emoions are not properly functioning. Emotions have therefore become more important in artificial intelligence, e.g. for improving rational decisions by reducing the search space, for increasing the acceptance of AI-programs by recognizing emotions of the user and by processing and expressing emotions, and by giving computergenerated actors (e.g. in edutainment or videogames) a “character” by developing albeit simplified personality models. In this presentation, we would like to first explain the motivation for this work, and then give an overview on how emotions are structured, their functionality analysed (or assumed) and, sometimes, already modelled, from different standpoints, e.g. the brain researcher, the cognitive scientist, the computer program and intaface developer, and the computer game developer. Finally, we will describe two potential impacts of emotions research, both on our self-view, namely the potential relation of emotions to consciousness and new ethical problems.

Emotionen für künstliche Akteure

Der Kybernetiker und AI-Forscher ROBERT TRAPPL untersucht menschliches Denken und die Moglichkeiten, es zu simulieren bzw. zu ubertreffen. Unter anderem mit Hilfe friedensbegabter Lernprogramme.