How can a student, lecturer, researcher or company employee
quickly access and compare the various terminologies, concepts, ideas
comments or advices contained in books, lecture materials, and
even "learning objects"? More generally, how can a person find
structured and precise information in answer to a precise question
(instead of being given a list of documents to read) and
how can a person publish some new idea or information without
having to write a whole document and thereby add to the volume
of "data" that other people will have to sort through?
Nowadays and still for quite a long time from now, the answer to the above questions
is "this cannot be done" because clearly no software is
knowledgeable and clever enough to second-guess people, understand the semantic content
of documents and merge it into a well structured semantic network.
Hence, people should store or retrieve concepts and ideas inside a
shared Knowledge Base (or inter-related KBs) and relate or retrieve
these concepts and ideas via semantic relations, especially transitive
relations such as specializationOf and subtaskOf.
Since 1999, I design the knowledge server WebKB-2 (www.webkb.org) to support this. During this project, I manually converted three courses into a semantic network and permitted the students of these courses to see, navigate and query this network. (No pedagogical strategy or student model was used to orient their navigation). This permitted them to better access and understand concepts and ideas, the descriptions of which were scattered across many documents. The students acknowledged the benefits of the approach but did not like having to learn the semantic network's notations and conventions. However, this is an unavoidable hurdle, as in any technical domain (e.g., sciences or classical music).
Besides supporting an "ideal" Learning Object repository, the tool and approach can (and ideally should) at the same time be used for permitting researchers to create semantically-structured cooperatively built states of the art of ideas and techniques, and for students or academics to give precise feedback on parts that need refinements. Students and researchers can also be evaluated via their contributions. Within companies, the approach could be used for precision-oriented corporate memories and would be a useful complement to Wikipedia. Details are provided in this article published during this project (www.webkb.org/doc/papers/elearn06/).