Modeling Ephraim Chambers' Knowledge Structure from a Naïve Standpoint

Abstract

In the preface to his Cyclopaedia published in 1728 Ephraim Chambers offers readers a systematized structure of his attempt to produce a universal repository of human knowledge. Divided into an interconnected taxonomic tree and domain vocabulary, this structure forms the basis of one effort from the Metadata Research Center to study historical ontologies. The knowledge structure is being encoded into a Simple Knowledge Organization System (SKOS) form as well as a Web Ontology Language (OWL) version. This paper explores the expressive and functional differences between these SKOS and OWL versions of Chambers’ knowledge structure. As part of this goal, the paper research focused on the construction and application of rules in each system to produce a more computationally ready representation of Chambers’ structure in SKOS, which is more thesaurus-like, and OWL, which represents additional ontological nuances. First, studying the various textual aspects at the semantic, syntactic, and typographic levels allowed for the relationships between terms to manifest from which rules governing expression of the connections between elements developed. Second, because each language, SKOS and OWL, functionally expresses different logical relationships, their possibilities and limitations offer a ground for further analyzing the resultant knowledge structures; although, each stemmed from the same basic source of Chambers’ text. Lastly this paper will examine rule making and expression in light of Paul Grice’s theory of conversational implicature to understand how a naïve agent formulates and applies these rules to a knowledge structure.