coMentor. A Collaborative WWW-Based Virtual Environment to Support Social Science Students — Graham Gibbs, Catherine Skinner, Andrew Teal

The coMentor project is producing a multi-user discussion system for the World Wide Web (WWW) which provides both a collaborative virtual environment where students can take part in discussions on theoretical issues related to the social sciences and humanities, and a set of software learning tools to support their debates. The project has developed software tools using HTML and Java to support access to a MOO (an object oriented multi-user discussion system) using standard WWW browsers such as Navigator and Internet Explorer. This provides a visual, cross-platform and familiar interface to students and allows the use of standard HTML features such as menus, buttons and text boxes to facilitate interaction and easy access to external resources. Enhancements to the learning environment are provided by Java applets some of which will support graphical "learning tools" such as concept mapping and structured argumentation tools, to support the students' learning processes.

Over recent years there have been many innovations in the application of information technology in the social sciences. However, use of information technology has been mostly in improving specific component tasks in the research process such as data analysis, coding of transcripts, word processing or presenting data and information. The development of hypermedia techniques, ie., hypertext and hot buttons creating linkages to sound, video clips and photograpic information has the potential to radically change the nature of the research process, facilitating the collection of a wider range of materials in the field. The use of hypertext means that it is possible to create multiple linkages and trails that that constitute complex and flexible pathways through the research material and between written papers and additional information. Thus hypermedia techniques can support new and exciting opportunities for more diverse and innovative representations of the research process. The incorporation of multi-media techniques in the research process can facilitate more open and transparent representations of research. Such representations can also reveal more of the often difficult process of analysis of raw material that is necessary to produce academic texts. This paper described the work about to be undertaken under an ESRC funded project to investigate the ways in which mult-media techniques can be used to represent an ethnography.

This paper reviews applications of the Internet and World Wide Web in the social sciences with reference to their graphical and visualisation content. Particular attention is given to recent developments in the use of on-line spatial decision support systems to improve public participation in social science research. Subsequent discussion focuses on (a) the (geo)graphical data requirements of such systems, (b) the potential future role of on-line participatory decision environments in reforming the democratic process, and (c) the social, cultural, political and ethical questions arising from such ideals.

This paper will briefly explore the contribution that computer modelling and animation now brings to landscape design teaching. Central to the work of landscape architecture is the graphic exploration and communication of both emerging ideas and completed design proposals. Developments in computer graphics have equipped the profession with a new set of tools that enable them to explore and present their ideas in a form that is more immediate to the designer and potentially more accessible to the non expert client/user. The paper will go on to identify how recent developments in internet technology including VRML, Java and Video Conferencing provide an opportunity to develop shared virtual environments which could be used for cross discipline collaboration in design teaching concerned with the built environment.

An exploratory spatial analysis tool known as MAPEX (MAP Explorer) has been developed; see Openshaw and Perree (1997). MAPEX creates MPEG movies of a search for map patterns by a set of smart pattern hunting creatures based on ideas from Artificial Life. The conjecture is that non-expert users may gain some insights into the types and locations of localised patterns in the GIS database by watching computer animations of the search process. The problem was that the complexity and inefficiency of the AVS component used to generate the MPEG movies was so cumbersome that all development of MAPEX ground to a halt. This paper describes the use of Java to provide an alternative animation tool that can be used to avoid the AVS bottleneck in MAPEX and similar GIS related animation needs.

Java offers a number of advantages: it is portable and platform independent, it is fast enough to allow user interaction and interrogation during the playback of stored map animation sequences, and it facilitates distributed experiments with the readers being able to perform their own visualisation experiments. The paper describes how the AVS component in MAPEX was replaced by Java and outlines various planned developments in Java based visualisations related to GIS and the exploration of high dimensionality geocyberspaces.

This paper suggests an application of Artificial Intelligence (AI) for the Social Sciences domain, which holds visualisation of past experiences. Knowledge in Social Sciences is poorly structured, making it difficult to develop computer algorithms to help the decision-making process. It is a domain highly dependent on experts, who generally take decisions intuitively, with a great deal of reliance on their judgement. It seems to be the perfect scenario for the application of the Case-Based Reasoning (CBR) paradigm. CBR holds representations of past experiences which can be reused to solve problems. Even in situations where no past experiences are available, CBR may help by providing a similar case from which users could base reasoning on and provide a solution. Moreover, continuous use of the CBR, expanding the case repository, can provide a corporate knowledge able to support tasks such as learning and training.

The following provide potentiall useful technologies: Java, VRML, AnswerWeb (some work at Edinburgh to add voice mail was noted).

There is a need for a central WWW resource for the social sciences — maybe what is needed is more information on SOSIG (not well known by participants) and other relevant services, such as the ESRC's QUALIDATA Service. We need to raise awareness of the potential of WWW related technologies, e.g. VRML, Java. CTI Centres could play an important role in promulgating techniques and tools.

The user can be the author of the end result and can be empowered by the provision of suitable visual interfaces to data. The interface can engage the user in collaborative work, for example in archaeology where a dig is discussed over the WWW and people collaborate in making decisions on where to dig next. There is a huge potential for creating collaborative decision environments on the WWW — intranet and more open WWW solutions.

Collaboration and open-ness may go against the traditions in some areas such as ethnography and we have to be sensitive to this. Copyright and confidentiality are big issues and go outside the scope of this event — but cannot be ignored.

These conclusions came out of 4 group discussion sessions each introduced by a paper. Groups were asked to focus on the questions highlighted below.