Stuart Watt

Artificial Societies and Psychological Agents

Agents have for a while been a key concept in artificial intelligence, but often all that the word refers to is a computational process or task with a capability for autonomous action, either alone or in an artificial society of similar agents. However, the artificial nature of these societies restricts the flexibility of agents to a point where social interaction between people and agents is blocked by significant social and psychological factors not usually considered in artificial intelligence research. This paper argues that to overcome these problems it will be necessary to return to the study of human psychology and interaction, and to introduce the concept of ‘psychological agents.’

Oracles, Bards, and Village Gossips, or Social Roles and Meta Knowledge Management

Knowledge management systems are used widely in many different organizations, yet there are few models and theories which can be used to help introduce and apply them successfully. In this paper, we analyze some of the more common problems for knowledge management systems. Using this background, we adapt models and theories from social and organizational psychology and computer supported collaborative work, and discuss a variety of different knowledge management systems in these contexts. We argue that knowledge management systems routinely adopt different social roles within an organization, and that these social roles can have a major influence on a systems acceptability. With these principles in mind, we draw out some general practical lessons, and a “character space” framework, which can help to inform the design of future knowledge management systems, so as to minimize the problems of acceptability within a given organization.

Role conflict in groupware

Groupware and computer-supported cooperative work (CSCW) systems are usually considered collaborative, but some kinds of group activity can actually be competitive. The author examines the differences between systems with one role and with several roles and how this potential conflict between the different roles arises. Design rules are provided to reduce the problems of this conflict. These rules are illustrated by an example of a large collaborative system involving many different roles.<<ETX>>

Learning by Building: A Visual Modelling Language for Psychology Students

Abstract Cognitive modelling involves building computational models of psychological theories in order to learn more about them, and is a major research area allied to psychology and artificial intelligence. The main problem is that few psychology students have previous programming experience. The course lecturer can avoid the problem by presenting the area only in general terms. This leaves the process of building and testing models, which is central to the methodology, an unknown. Alternatively, students can be introduced to one of the existing cognitive modelling languages, though this can easily be overwhelming, hindering rather than helping their understanding. Our solution was to design and build a programming language for the intended population. The result is Hank, a visual cognitive modelling language for the psychologist. Our informal analyses have investigated the effectiveness of Hank in its intended context of use, both as a paper and pencil exercise for individuals, and as a computer-based project to be carried out in groups. The findings largely support the Hank design decisions, and illuminate many of the challenges inherent in designing a programming language for an educational purpose.

Hank: a friendly cognitive modelling language for psychology students

Cognitive modelling involves building computational models of psychological theories in order to learn more about them, and is a major research area allied to psychology and artificial intelligence. The main problem is that few psychology students have previous programming experience. The course lecturer can avoid the problem by presenting the area only in general terms. This leaves the programming side, which is central to the methodology, a closed book. Alternatively, students can be introduced to one of the existing cognitive modelling languages, though this can easily be overwhelming, hindering rather than helping their understanding. Our solution was to design and build a programming language for the intended population. The result is Hank, a visual cognitive modelling language for the non-programmer. The design of Hank drew on research areas including end-user programming and software visualization to find an appropriate design. A number of empirical studies are planned.

Managing Text Objectively

Complex development environments which use languages of diverse styles and purposes have special problems efficiently managing their data. Generally there are two alternatives; first, marking the text with extra codes, and second, using a tool capable of structuring the data, such as an object management system. For many purposes, something between the two is appropriate, in which plain text files can be read and used as if stored in an object management system. This paper describes a technique which can be used to implement this kind of data management substrate, and how it can be used to help in the construction of one such complex environment, the Vital workbench.