Re: "Databases are Categories"

["Dr. Cyrus F Nourani" <acdmkrd@gmail.com>]

Hello, thought I state the "obvious"?  since I had written an ACM review for
newer techniques to
model database and program transformation techniques a couple? years ago.
There were papers by our UCLA Semantics and TCS group two decades ago with
publications on database algebras. Further specificson databases and
algebras
were at TU Berlin.

I enclose an excerpt:

[There are publications at UCLA,JACM, and TU Berlin since 1979 that address
abstract algorithms, programs, morphisms, and database models and schemas
that are benefit the areas.There are  mapping techniques that can be applied
to the data processing realms: There is a concrete syntax route, e.g.,
untyped canonical mappings sparse trees using language independent
format-universal types,  generative mappings- metaprogram- GDK grammar
deployment GUI, XML data biding, and the problem of providing an object
model that is to represent an XML schema are stated. Graph Transformation
for Model Refactoring deploys models and transformations as primary
aritifcats. The techniques presented are to use graph model representations
and apply graph transformations at the model refactoring arena.
Refactoring (Opdyke) is changes to the internal program structure to improve
without changing the external introduce model refraction as a new
transformation. The Berlin school AGG is applied on graph grammars and to
specify model refactoring.]

CyrusFN

-- 
Akdmkrd.tripod.com
Projektakdmkrd.tripod.com


On Sun, Aug 15, 2010 at 11:25 AM, Pym, Professor David J. <
d.j.pym@abdn.ac.uk> wrote:

> John Cartmell (who introduced contextual categories and generalized
> algebraic theories as models of dependent types in his 1978 Oxford D.Phil
> thesis and a 1986 Ann. Pure App. Logic paper) worked on topics related to
> this thread some years ago. Another 1986 paper, 'Formalizing the Network and
> Hierarchical Data Models --- an Application of Categorical Logic', CATEGORY
> THEORY  AND COMPUTER PROGRAMMING
> LNCS, 1986, Volume 240/1986, 466-492, DOI: 10.1007/3-540-17162-2_138, can
> be found via the following
> link:
>
>         http://www.springerlink.com/content/y31234tkk63wp56k/
>
> 'Abstract. We have noted that data modelling and conceptual modelling have
> content and performance as their concerns. For the different data models,
> the Network and the Hierarchic, we have given logics involving the
> operations which are physically supported according to the data model. The
> logics are sensitive to performance in a way that classical logic is not. We
> have now suggested how we might formalise this. Network and Hierarchical
> databases  have the functional inverse or family as their primitive of
> organisation. To formalise the Network model we have given a general
> definition of network category which seems to generalise correctly the
> hierarchical logic of contextual categories.'
>
> There is also relevant work on categories and logic programming.
>
> David Pym
>

[For admin and other information see: http://www.mta.ca/~cat-dist/ ]