Testing a Virtual Campus Prototype
Claude RICCIARDI RIGAULT <firstname.lastname@example.org>
This paper presents a virtual campus delivery prototype case, tested in a real-life telelearning situation, with students disseminated across the state. The learning environment integrated 18 computer tools facilitating communication, information gathering, learning management, and navigation. It was supported through a heterogeneous network including ATM, ADSL, and modem cable, with the Internet Protocol. Last year, we tried a similar pedagogical scenario using an ISDN network.
The actors represent any entity interacting or exchanging information in the system, be they people or pieces of software. In the act of learning, we distinguish five actors categories: a learner who wants information to transform it into knowledge, an information provider, and three facilitators in the information conversion process, such as a designer who makes information easy to absorb, a trainer who assists the learner in the conversion process, and a manager who provides an easy context to the process.
Each one of an actor's roles can be seen as a group of processes that an actor can execute. Each process is a set of actions that a user can perform while interacting within the virtual campus. A process is one of the ways to use the system. The whole set of available processes define what the system is used for and demand a specific configuration.
The underlying architecture includes five spaces for:
The Learning Scenario of a Learning Event is a graph composed of all possible paths between its constitutive learning subevents. Adding to the paths are the didactical media and production tools necessary to execute a Learning Event. Special attention was given to the possibility of integrating any kind of document or application inside the Learning Scenario through the concept of reference.
The Learning Scenario defines the default learning path proposed to a learner for a Learning Event established by the designer. A Learning Scenario may be adapted by a learner (or a team of learners) to become a personalized Learning Scenario.
The basic material used to build the Learning Scenario model is a generic foundation called the Generic Graph. Not directly accessible to the actors, it constitutes the basis of all the knowledge, scenario, media, tools, collaboration or assistance data, together with their interrelations. The Generic Graph is the fundamental structure from which every component of an editor is derived. The Virtual Campus includes at least three types of editors: the knowledge model editor, the learning model editor and the media editor. These editors feed corresponding navigators on the Web.
This general TeleLearning model has been implemented in a prototype, taking into account all of the above aspects with regard to the learner's and the trainer's configurations. It was not the first prototype conceived by the LICEF's researchers and took place in an iterative process of developing the Virtual Campus. It was tested over seven weeks by 15 students, each working at a distance from the others, in June and July 1997. Students received help from two counselors.
It consisted of a reengineering of a telecollaborative method used in a previous course offered by the Tele-Universite. The treated subject matter was the relationship between continuing education in the workplace and competitiveness of the enterprises, presented through three models, namely the German, Japanese and Swedish ones. The Learning Scenario favored collaborative work, preset or spontaneous, either in synchronous or asynchronous mode. Activities included four processes: delineation of a problem, information retrieval, individual proposals, and collective debate. A fifth process, socialization, tagged along with the others. Learners formed into teams of five members, where the different points of view contributed to hypothetical solutions, which in turn were examined and criticized by the other teams. Finally, a spokesperson for each team participated in a general debate.
The five spaces of the Virtual Campus were presented to the learners by the way of an interface called Explora, assisting the learners' navigation. This first interface was based on the metaphor of a network, each space being represented by a constellation. Most of the 18 different tools offered to the users appeared by double-clicking on one of the planets. On the other hand, the specific course environment was built around another kind of spatial metaphor, a Virtual Congress Center, where spaces symbolized the different processes necessary to achieve the Learning Scenario.
The tools have either been home-designed and -developed, or simply imported and integrated in the global design. We may categorize them in terms of the five spaces mentioned above, that is to say: information, production, collaboration, assistance, navigation and self-management. We may also present them by the way of the following figure:
We experimented with a heterogeneous type of network. That is to say, seven learners were supported on ATM and four on ADSL while four others were linked by cable-modem. Personal computers were all Pentiums with an adequate and defined configuration. Learners could explore the same functionalities but, depending on the type of network they were connected to, not always by using the same specific applications. For example, in the case of videoconferencing, some used the multipoint multimedia videoconferencing system called MMOne from General DataComm while others were using Connectix at the same time and on the same object. One of our purposes was to try to combine those tools and to compare them in relationship with the kind of network used. During the summer of 1996, we had been able to test a similar setup on ISDN, so we now have a good idea of the range of available means to deliver TeleLearning systems on networks.
The first challenge addressed our ability to integrate so many tools, issued from diverse areas, into a coherent, friendly and efficient environment.
The second was to make people -- learners and trainers -- able to manage such a range of electronic tools without disturbing specific matter learning.
The third was to combine a two-level experiment: our own complex learning and technical setup experiment and another one in which some of our partners were experimenting on the network (ADSL, for example) or piece of software through which it was conducted.
Specific challenges, among others, lay in using multipoint technology for interaction specific to collaborative work; another point consisted in offering an information system that allowed ulterior intelligent manipulation of documents, in order to enable learners to build knowledge bases.
At the collaborative learning level, we shall specifically examine means of teleconferencing. Teleconferencing features group discussion, exchange and work on given topics. Participants learn by progressively building and elaborating on content together. However, to be an efficient tool in distance learning, teleconferencing poses some conditions. Experience taught us that two principles must be followed:
The content analysis of several teleconferences led us to identify at least four types of messages:
There must be a particular site designed to accommodate each type of message. In the participants' eyes, the teleconference that is task-oriented seems to be the most important. However, the planning and design of other types of teleconferences should not be overlooked, in order to avoid an information overload within the task-oriented one. Otherwise, the absence of other vocational sites will negatively affect the group performance and possibly the learning potential. Each virtual space should have its own moderator to support and regulate the debates. In the context of a collaborative learning situation, the moderator's role is that of a facilitator.
The special tool developed for textual information processing, information retrieval and assistance to knowledge-based building has had great success. The reason for that success is revealed in its uses. FX Workshop (also called Nomino) makes it possible to develop knowledge bases using natural language as an information support; it helps information retrieval in a natural way; it gathers information of the same kind; it enables multiple points of view on the same material; it helps to detect convergence; and it supports comparison, diagnosis, etc.
The same tools can be used to perform search and information retrieval as well as textual analysis, to assist in hypertextual reading, to elaborate upon multi-indexing and annotation cards (called e-cards), and to help in categorization -- predetermined or emergent. In a formal way, it is akin to the elaboration of a database; however, the kind of structure used helps to easily multiply the categories. Some features also permit an elaboration or change to the categorization scheme without discarding the work already done. One of the students told us that, thanks to the pedagogical scenario and the Nomino tool, for the first time she really felt that she was building her own knowledge.
The FX-Net software is used to program dynamic links between the Internet and the electronic workshop. Each e-card may contain an Internet source (integrating a clip or associating the complete text site to the e-card). It is also possible to give the e-card a categorization scheme, building a bookmark base at the same time, allowing easy further exploration by the way of the Nomino engine. It is possible to delegate to an FX-Net agent the task of exploring a Web site and to automatically build new e-cards based upon it.
FX-Net integrates automatic Web site building features by using, on one side, the e-cards linked to their hypertextual sources and, on the other side, the index classifying the cards by categories. It is also possible to dynamically feed a Web site by way of the e-card bases. After applying the Nomino parser on a text, the workshop can also automatically build an HTML page, composed of the full-source text, indexes created by Nomino (allowing navigation inside the text), and a 'digest' of the original text.
In terms of technical issues, overall findings reveal that students in this high-tech environment rely on synchronous communication when in need of help or motivation but prefer asynchronous communication to learn content. Moreover, it is evident that in order to appropriately serve students and tutors at a distance, the technological expertise must be both varied (telecommunication, hardware and software) and distributed. In terms of pedagogical issues, students felt a little overwhelmed by the plethora of tools that they had to learn to manipulate, which in turn had a negative effect on the academic learning in the first weeks. Methodologically speaking, it was found that students need a goal-oriented approach to learn and to appreciate new technology.