Use this URL to cite or link to this record in EThOS:
Title: Exploratory discovery : a content theory of self-questioning and experimentation
Author: Oehlmann, R.
Awarding Body: University of Aberdeen
Current Institution: University of Aberdeen
Date of Award: 1996
Availability of Full Text:
Access from EThOS:
This dissertation describes a novel computational model of an autonomous agent capable of learning new explanations. The model views learning as a goal-based process which can be achieved by planning the generation of questions, answers, and experiments. This view is characterised by a set of interactions between the processes of generating questions and experiments. The interactions are based on a novel content theory which models the exploration of alternative paths of reasoning and acting. Therefore the content theory is the main focus of the thesis and includes a taxonomy of typical reasoning situations in which questions can be used. Moreover, the content theory supports learning in unfamiliar domains; we therefore refer to the computational model as Exploratory Discovery. The model has been implemented in the IULIAN system which uses case-based reasoning and case-based planning approaches. Whereas the reasoning approach enables the system to re-use its knowledge about previous experimental results, the planning approach supports the reuse of previous plans for generating questions, answers, and experiments. The IULIAN system has been used to investigate the Exploratory Discovery model in the domain of electric circuits, water pipes, and the chime of a mechanical clock.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available