Brian R. Gaines and Mildred L. G. Shaw
Knowledge
Science Institute
University of Calgary
Alberta, Canada T2N
1N4
{gaines, mildred}@cpsc.ucalgary.ca
Concept mapping has a history of use in many disciplines as a formal or semi-formal diagramming technique. Concept maps have an abstract structure as typed hypergraphs, and computer support for concept mapping can associate visual attributes with node types to provide an attractive and consistent appearance. Computer support can also provide interactive interfaces allowing arbitrary actions to be associated with nodes such as hypermedia links to other maps and documents. This article describes a general concept mapping system that is open architecture for integration with other systems, scriptable to support arbitrary interactions and computations, and cutomizable to emulate many styles of map. The system supports collaborative development of concept maps across local area and wide area networks, and integrates with World-Wide Web in both client helper and server gateway roles. A number of applications are illustrated ranging through education, artificial intelligence, active documents, hypermedia indexing and concurrent engineering. It is proposed that concept maps be regarded as basic components of any hypermedia system, complementing text and images with formal and semi-formal active diagrams.
Hypertext systems commenced with simple facilities for creating nonlinear texts by embedding within a text links to other texts (McKnight, Dillon and Richardson, 1991). Hypermedia systems are a natural evolution of the linked content notion in which links can be embedded in any presentable item, and connect to any other one (Barrett, 1989). In particular, diagrams become hypermedia objects with the possibility of links from text into diagrams, and from within diagrams to other diagrams, text and general hypermedia objects. Concept maps are a form of diagram specifically targeted to provide visual languages similar in their characteristics to natural language text in that they can be subject to syntactic and semantic constraints, and their representational capacity can range from the fairly informal to the extremely formal.
Concept maps have been used in education, policy studies and the philosophy of science to provide a visual representation of knowledge structures and argument forms. They provide a complementary alternative to natural language as a means of communicating knowledge. In many disciplines various forms of concept map are already used as formal knowledge representation systems, for example: semantic networks in artificial intelligence, bond graphs in mechanical and electrical engineering, CPM and PERT charts in operations research, Petri nets in communications, and category graphs in mathematics.
Support of a general concept mapping component is appropriate in the architecture of any hypermedia system. The work reported in this article has originated from the development of systems to support large volumes of heterogeneous multimedia material generated in a variety of contexts such as knowledge acquisition (Gaines and Shaw, 1992b), sports coaching (Vickers and Gaines, 1988) and large-scale project support (Gaines and Norrie, 1994). The primary requirement has been to give end users access to multimedia material through indexing mechanisms that are simple and natural to use. A further requirement has been to support shared access to material through a variety of interfaces that reuse the same material for different purposes (Kremer and Gaines, 1994).
These requirements have been addressed by the development of a general visual language technology supporting customizable interactive concept maps (Gaines and Shaw, 1993c) and semantic networks (Gaines, 1991). The maps may be used as stand-alone documents or embedded as interactive pictures in active documents (Gaines and Shaw, 1993b). The technology is open architecture, and user interaction with the concept maps may be programmed to initiate any activity feasible on the host system. Thus maps may be linked to other maps for retrieval purposes, and may be used to retrieve, play and edit multimedia material either through the host system or under its control. The technology is designed to be shared and operate over local or wide area networks so that multiple users can access the same maps, and maps can be used to retrieve material from remote sites.
Concept maps may be used for the indexing and retrieval of hypermedia material, providing an attractive, meaningful and easy to use interface. In addition, maps with formal semantics may be used to provide a programming interface for the control of multimedia material. For example, Petrinets may be used to specify the essential synchronizations in the playback of a mix of multimedia material, and to index through the time relations in such material. Semantic networks may be used to represent knowledge bases that are operational and can be used for knowledge-based access to material.
The following sections give an overview of concept maps and their applications, formal semantics for concept maps, an abstract model of general concept maps, the implementation of a general concept mapping component for hypermedia systems, its user interface and computational capabilities, its application to hypermedia indexing, collaborative applications of linked maps, and integration of concepts map with World-Wide Web systems for wide-area collaboration.
Many disciplines developed visual languages for `concept maps', `cognitive maps' and `argument forms.' In education, Ausubel's (1968) cognitive learning theory led Novak (1977) to develop a system of concept maps that has been widely applied in the evaluation of students' learning in the school system (Novak and Gowin, 1984). Figure 1 shows a concept map from these studies with two types of nodes, concepts shown by ovals, and instances shown by rectangles. These are linked by arrows labeled with relations such as needed by, made of, changes, and so on. The conceptual structure developed encompasses some of the physics and biological roles of water. The student has developed the map within broad guidelines as to what are concepts and instances, and that they are to be linked by labeled directed arrows denoting relations.
Figure 1 Concept map of student's knowledge (after Novak and Gowin, 1984)
A wide variety of different forms of concept map have been applied in education (Lambiotte, Dansereau, Cross and Reynolds, 1989). They have also been used as tools to support the interviewing process in knowledge acquisition from experts, for example in the Wright-Patterson development of the pilot's associate (McNeese, Zaff, Peio, Snyder, Duncan and McFarren, 1990). In management, Axelrod (1976) proposed a form of concept maps as a means of representing the conceptual structures underlying decision making, and these have been used empirically to analyze organizational decision making (Eden, Jones and Sims, 1979), social systems (Banathy, 1991) and the policies of political leaders (Hart, 1977).
In artificial intelligence, Quillian (1968) developed a form of concept map that came to be termed semantic networks and used extensively for formal knowledge representation. In linguistics, Graesser and Clark (1985) have developed an analysis of argument forms in text in terms of structured concept maps with eight node types and four link types, and Woodward (1990) has developed tools to extract such maps from text. In the history of science, the dynamics of concept maps have been used to represent the processes of conceptual change in scientific revolutions (Nersessian, 1989; Thadgard, 1992).
In the philosophy of science, Toulmin (1958) developed a theory of scientific argument based on typed concept maps that is regarded as one of the major themes of the rhetoric of western thought (Golden, Berquist and Coleman, 1976). Figure 2 shows a concept map from these studies applied to the derivation that Harry, born in Bermuda, is a British citizen. The methodology prescribes various types of component that may be expected in argument forms such as backing, warrant, and so on, and these are indicated by textual labels rather than differing shapes.
Figure 2 Concept map of argument form (after Toulmin, 1958)
Hypertext systems that display overviews of nodes and links provide a natural tool for the computer support of concept map development (Smolensky, Bell, Fox, King and Lewis, 1987), and specific tools to do so have been developed (Streitz, Hannemann and Thüring, 1989; Bernstein, 1992). Frames, semantic networks and hypertext structures have been compared (Travers, 1989), and Jonassen (1990) has shown how the elicitation of semantic networks may be used to structure hypertext. A variety of techniques and computer-based tools have also been developed for the analysis of concept maps. When they represent the strict semantics of a visual language for knowledge representation, it is appropriate to use deductive inference engines to draw logical conclusions from the maps (Sowa, 1991a). Even general concept maps can be analyzed in terms of their structure of their link relations (Cropper, Eden and Ackermann, 1990).
gaines@cpsc.ucalgary.ca 22-Nov-95