Use this URL to cite or link to this record in EThOS:
Title: Knowledge mentoring as a framework for designing computer-based agents for supporting musical composition learning
Author: Cook, John
ISNI:       0000 0000 3885 4707
Awarding Body: Open University
Current Institution: Open University
Date of Award: 1998
Availability of Full Text:
Access from EThOS:
Access from Institution:
An approach to the design of teaching agents in problem-seeking domains - that is based on a systematic relationship between theoretical framework, analysis of empirical data, computational model and computational implementation - has been developed. The theoretical framework, called the Knowledge Mentoring framework (KMf), was developed to investigate how studies of dialogue and interaction can be exploited in a practical way by designers of computer-based teaching agents. A particular focus was the following musical education problem: when interacting with a computer-based music system, many students do not spontaneously reflect on their activity, they often need to be encouraged to do this. The KMf provides a taxonomy and definitions of the pedagogical goals involved in a 'mentoring' style of teaching. Mentoring is an approach to teaching that aims to support learners' creative, metacognitive and critical thinking, these being essential to musical composition and other open-ended, problem-seeking domains. This theoretical framework was used to guide the analysis and modelling of data produced by an empirical study of human teacher-learner interactions. Information on the temporal ordering of teacher-learner interactions was revealed (modelled as. state transition networks and a mentoring script). Findings from the analysis also included a pause taxonomy (that provided evidence of a link between pause length and learner ability) and the occurrence of reciprocal modelling (where participants in learning interactions built up models of the other participants' expectations). The theoretical framework and the analysis findings were then used to develop a computational model for teaching agents in problem-seeking domains. Aspects of our theory, analysis findings and computational model were incorporated into a computational implementation: a pre-prototype teaching agent called MetaMuse. A Cooperative Evaluation of MetaMuse with teacher-composers showed that it had the potential to promote creative reflection in learners.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: Metacognition; Creativity