Future Directions ... or at least something to think about

[bugmagnet]

G'day everyone,

My colleague Diarmuid Pigott has just had his PhD thesis published. It can be found on the Murdoch University website.

The blurb on the website is below. How Oxygen Basic might benefit, I don't know, but it's something new in the knowledge modelling space and well worth a good hard look.

Kind regards,
Bruce/bugmagnet.

Conceptual modelling of knowledge has remained an open research challenge. Existing frameworks do not cope with problems such as multiple user viewpoints, the plurality of epistemologies and representational forms, the mutability of knowledge, and the great body of legacy encoded knowledge. This thesis addresses the lack of a systematic method for the conceptual modelling of knowledge by presenting a novel perspective for dynamic knowledge exchange together with an associated modelling framework and tools.

The thesis establishes a new perspective, the erotetic perspective, based on question-and-answer exchanges that match knowledge needs with knowledge capacities. It presents a unified design framework within this perspective and introduces appropriate modelling constructs, the Functional Entity and the Knowledge Relation. The framework comprises a methodology (the Functional Entity Relationship Methodology), a diagramming system for drawing conceptual models (the Functional Entity Relationship Diagram) and a transactioning language for representing the knowledge exchanges (the Functional Entity Relationship Language). These respectively extend the classic Entity Relationship Diagram and the class of SQL-like languages adequately to describe all possible transactions involving encoded knowledge. The different types of Functional Entity are shown to cover the complete space of knowledge seeking and retrieval and cope with situations not possible in conventional data modelling.

As the modelling framework is a secondary design artefact (one that is capable of producing routine design artefacts) the design science research approach of Gregor & Jones was used. This approach necessitates an Alexander pattern drawn from prior research to guide development, followed by expository instantiations of the artefacts sought. Evaluation comprising verification, validation, generalization, substantiation and some external accreditation was conducted throughout. The models developed were tested for mutual encompassing through docking, which also confirms the erotetic perspective. Illustrative cases are presented to show the completed framework in action.

[Charles Pegge]

Hi Bruce,

I had a quick scan of the thesis. Hard to grasp yet fascinating. From a developers point of view, we tend to work bottom-up. (I started Oxygen as a machine code script), so building 'Mount Abstract' takes a while.

I think a highly standardised OOP based framework would do the job nicely, using well defined objects/nouns and verbs/methods, and message exchange protocols.

Multi-Actor games, probably have much of the required logical framework already.

Charles