The University of Arizona

Data Cognition


In London, at the very moment this is being written, the twenty-odd members of the International Standards Organization SQL working group are meeting to deliberate proposals for changes to and extensions of the SQL3 standard. Some of these proposals are minor tweaks and some represent significant additions to the language. The same vexing questions from prior discussions over the years have arisen yet again. Should redundancy or orthogonality of constructs be preferred? Should the language changes be based primarily on a declarative (i.e., predicate calculus) orientation or should a procedural (i.e., algebraic) orientation also be considered? At what point does language complexity become an overriding consideration?

Project Overview

Surprisingly, there is no detailed model that answers this question. A review of the data management literature that has blossomed over the last 40 years reveals a fragmentary, often contradictory, and incomplete picture of how people organize and utilize data.

This project is a concerted, multi-faceted effort to develop an underlying basis for data management based on the articulation and scientific validation of a theory-based user's mental model of data semantics and of data access. This model can be helpful in evaluating existing conceptual modeling and query languages and for designing more effective languages.

Richard Snodgrass (University of Arizona)
Glenn Browne (University of Virginia)
Vijay Khatri (Indiana University)
Jeffrey Parsons (Memorial University of Newfoundland, Canada)
Christian S. Jensen (Aalborg University, Denmark)

Webmaster: Sam Martin