Knowledge Encoding and Retrieval Tool


Brian Matthew Winn

Faculty Directors:
Dr. Carrie Heeter
Dr. Don Weinshank
Dr. Bill Punch

November 8, 1994


Hypermedia, also known as interactive multimedia, has been heralded as the next-generation learning tool. Hypermedia systems have the ability to incorporate large bodies of information with multiple forms of media representation (text, graphics, audio, video). The real power of hypermedia as a learning tool does not lie in the wealth of information that any one system can contain (that is a database problem), rather, it lies in the potential of the system to give the user access to the right; information at the right; time and the potential of the system to present the information to the user in a representation that motivates the user and promotes learning. Unfortunately, these two capabilities are often missing and have yet to be fully realized in hypermedia systems today.

The Knowledge Encoding and Retrieval Tool (KERT) project is an attempt to address these two challenges in hypermedia system design. The goals of the project are threefold: a literature review, the construction of a prototype, and the construction of the knowledge encoding and retrieval tools.

The literature review will concentrate on the areas of artificial intelligence, hypermedia, and education. 1.) The hypermedia lit review will focus on the early work with Hypertext such as Brown University's"FRESS" and "Intermedia" systems and Ted Nelson's "XANADU" extending to current work which tie in multimedia elements such as Interval Research's "Media Streams" 2.) Learning aspects of the lit review will focus on learning through hypermedia, particularly hypertext, including some basic research on the most relevant pedagogy theories such as Papert's situated knowledge and constructivism. 3.) The AI literature to be searched will focus on goal-directed search and discovery of information via an intelligent agent. Literature will be sought and reviewed that identifies knowledge encoding methods to apply to the actual media data. These encoding methods could range from a simplistic keyword matching to a more complex pattern matching rule-based type system where rules would match patterns of text to concepts. More complex natural language approaches will also be investigated. A report will be generated summarizing the findings in the literature review.

This research hopes to identify methods of encoding information for retrieval by intelligent agents, methods of visualizing hyperlinks in large and potentially diverse domains and navigating through them, and methods of constructing effective hypermedia learning tools to enhance the hypermedia experience. The hypermedia prototype will experiment with various information encoding and multimedia techniques and is meant as a proof of concept. The prototype will concentrate on a specific domain of knowledge, most likely scientific, and will be meant as an educational learning and discovery tool in that domain. It will most likely be based on one of the knowledge bases under development at the Comm Tech Lab.

The prototype will be constructed in Kaleida's new ScriptX multimedia authoring environment. ScriptX, which is currently in beta testing, was chosen because of its allegedly powerful object-oriented scripting language, its cross-platform portability, and its growing market support. If ScriptX is unavailable or is deemed inappropriate or insufficient for the job, other development environments, such as Apple Dylan, will be investigated.

The final phase of the project will entail the construction of a toolbox of selected KERT objects. This toolbox will allow future hypermedia projects to quickly and easily incorporate the ideas developed in the prototype.

Applicability of Project

This project strongly relates to the interests of the Comm Tech Lab. The CTL hopes to incorporate the ideas developed and researched, as well as the toolbox of KERT methods into many of their future multimedia projects. One of the first projects that KERT is hoped to have a direct impact on is the
(c) 1995 Communication Technology Laboratory
Michigan State University