The initial motivation behind the Knowledge Web was to enhance Aristotle,
an online classroom of Rutgers University. Aristotle [2]
is a web-based educational tool that supplements classroom instruction for a
Rutgers University freshmen course in general biology. It contains course
material divided into lectures, which in turn are organized as topics and
keywords. Each topic/keyword delivers information organized in the form of a
multimedia presentation. The presentation integrates a piece of video,
corresponding transcript, definition and images, all relevant to the particular
topic. The presentation is delivered in a browser that embeds a RealPlayer(TM)
plugin. It can be rewound, fast-forwarded or paused. This integrated way of
presenting course content to students was very well received. The rendering
of lecture videos into topics and keywords enabled students to see just that
portion that pertained to their topic of interest.
Despite its success and uniqueness as an online classroom, Aristotle cannot
function as a replacement for the traditional classroom. It lacks the
interactivity present in a student-teacher classroom and does not contain a
substitute for a good teacher. Aristotle has course content organized into
lectures and the lecture index pages contain hyperlinks to presentations of
topics and keywords. Although this communicates a lot of information, it does
not contain a way of representing relationships between topics. In order to
thoroughly understand a new topic, students need to compare and contrast it to
something they have previously learnt. Aristotle does not provide any means
to do this. Thus in order for Aristotle to function as an independent
classroom, it is necessary to extend it. This resulted in the research that
lead to the Knowledge Web.
The Knowledge Web does not belong to any traditional knowledge representation
category [3](KR). A typical KR method comprises of
two parts, automatic assimilation of information and graphically
representing them. They concenterate more on assimilating and constructing a
knowledge domain, rather than actually portraying them pictorially. Also, most
of these representation schemes are very specific to a certain subject. For
example, there is Grasp[4], a molecular visualization and
analysis program, particularly useful for the display and manipulation of the
surfaces of molecules and their electrostatic properties. Similarly, there
are a number of tools for computer network visualization and
management [5]. These tools are very specific to their
domain of interest and cannot be used to represent information in other
realms.
The Knowledge Web relates to the depiction of information or knowledge. It
graphically represents the information that has been collected and categorized
apriori. With the wide variety of information now available on the Internet,
information or knowledge depiction by itself forms a sizeable problem.
Information now appears in several different incarnations - images, videos,
animations, vitual reality simulations, plain text, HTML, powerpoint slides,
and many other forms of multimedia. Currently, there are no tools in the
market integrate all these information and depict them graphically. The
Knowledge Web solves this problem. It integrates varied pieces of information
relevant to a certain topic and depicts them in the form of a graph. A school
teacher may use the knowledge web to effectively teach her students a new
topic. The knowledge web also enables the teacher to depict relationships
between different topics. Added to all these advantages, the knowledge web is
saved as an XML document [6]. Hence a knowledge web that is
published on the Internet, can be easily searched by current search engines
for answers to specific queries.
The Knowledge Web may be compared to Cognitive maps [7]. A
cognitive map is a graphical representation technique used for analysing
knowledge. It is a realisation of a likely mental device that codes and
simplifies the way our spatial environment is arranged [8].
The Knowledge Web is designed on the same lines as a cognitive map. It
represents concepts and their relationships in the same way that cognitive
maps are constructed in human minds. Given this similarity, the information
presented by the Knowledge Web is easily understood, thus fostering
meta-cognition.
The remaining part of the paper explains details of the knowledge web, and the
way it has been implemented. Section 2 explains the concept.
Section 3 explain the graphics framework used to implement the
knowledge web applications, section 4 explains how the
knowledge web has been represented by means of an XML document, section 5 presents details of the applications that have been designed and implemented to construct and navigate through the knowledge web. We conclude the paper in section 6.