Classification Networks: A Knowledge Representation Scheme for Curriculum Prescription
Abstract
This paper presents a classification scheme that articulates the categorisation of subject matter. Two basic classificatory devices are introduced: (a) a feature representing the resulting category characterised by a property that the feature denotes, (b) a dimension representing a perspective that partitions the Universe of Discourse (UoD) or a category. A classification based on a dimension can be further classified into the categories it has formed. Dimensions can be juxtaposed to form a classification based on multiple perspectives. The classificatory devices of features and dimensions, which form a classification network, support a wide range of subject organisation types, viz. (a) conceptual organisation, (b) procedural organisation and (c) theoretical organisation. The purpose of the approach is to support the clarity, simplicity and maintainability of large scale general ITS development. Some simple course organisation examples using classification networks are also presented. 1. Background In the intelligent tutoring systems (ITS) community, there have been very few discussions about methodological issues such as the kinds of knowledge that can be used for what design purposes. In particular, when dealing with larger scale ITS developments, there is a tendency to confuse two different design processes: (1) prescribing the curriculum (what should be there for pedagogical purposes?) and (2) representing content knowledge (what should be there for the running of the tutoring system?). Curriculum prescription is a human-based design process, specifying what is important and which material is more important than the other. The (partial) ordering of the subject categories is subject to the emphases of the course and the student background. The representation of content knowledge is also a design process, but it is based on both human understanding and machine efficiency (epistemological adequacy). From the ITS designer's point of view, curriculum prescription and content knowledge
Cite
Text
Yum and Richards. "Classification Networks: A Knowledge Representation Scheme for Curriculum Prescription." International Joint Conference on Artificial Intelligence, 1993.Markdown
[Yum and Richards. "Classification Networks: A Knowledge Representation Scheme for Curriculum Prescription." International Joint Conference on Artificial Intelligence, 1993.](https://mlanthology.org/ijcai/1993/yum1993ijcai-classification/)BibTeX
@inproceedings{yum1993ijcai-classification,
title = {{Classification Networks: A Knowledge Representation Scheme for Curriculum Prescription}},
author = {Yum, Kwok-Keung and Richards, Thomas J.},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {1993},
pages = {447-452},
url = {https://mlanthology.org/ijcai/1993/yum1993ijcai-classification/}
}