Talk:Schema

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Stub article[edit]

From the main page:

this is a stub article - the idea of a schema needs a broader treatment than just being called a subset of an ontology

Yes. I agree, if only because I don't understand what that means. But I don't think this note should be on the page itself. --KQ 21:13 Aug 21, 2002 (PDT)

Genetic algorightms similarity schema[edit]

Is it a set of entities or the template that represents that set? -Pgan002 00:35, 19 October 2006 (UTC)[reply]

Database schema[edit]

Why are "software architecture" and "Active directory" mentioned in the disambiguation page?! This clearly violates the WP Syle Guide of disambiguation pages. I also think that "conceptual graph", "semantic network", "semantic web" and "data model" sould not be mentioned. Disambiguation pages should only have a brief definition of each use and refer the reader to a specific page. If there is no page about a specific use of a term, I think that either one should be created or the entry deleted. -Pgan002 00:43, 19 October 2006 (UTC)[reply]

Prior knowledge[edit]

Surely "schema" never means "prior knowledge"! -Pgan002 00:44, 19 October 2006 (UTC)[reply]

Specialized ontology[edit]

Is it the set of questions or the actual ontology, i.e. the entities?

Loose and unspecific?[edit]

As it stands, this (stub) article begins with "A schema in general is a specific, well-documented, and consistent plan. The related word, scheme means a loosely described plan." I don't know where this comes from, or why I should doubt my dictionary, but the statement implies that a scheme isn't specific, well documented, or consistent. What nonsense! And though they're both called "plans," schemes are loosely-described? If that's the case, all it would take is for somebody to describe a scheme "tightly" (whatever that means), and it would change from being a scheme to a schema! The fact is, "scheme" and "schema" both come from the same Greek word, and the latter (presumably more pedantic, since it's Greek) has crept into usage only to distinguish certain things from schemes, for whatever reason. My dictionary gives two meanings for schema: first that it's a scheme, and second, that it's a word used in Kantian epistemology. I know very little about Kant or epistemology, but I suspect that Kant wrote in Germanneologism, for the sake of cant! Unfree (talk) 08:01, 24 December 2007 (UTC)[reply]

Computing usage[edit]

In computing usage, I'd say that "schema" means the definition of a data structure - what its components are, and how they relate to each other. Database schemas, XML schemas, JSON schemas, and so on are all specializations of this concept to particular ways to expression data. None of the articles seem to present this unifying concept, though Data model comes close.

I'm also a little surprised that there are no mentions anywhere of metaschemas, which are the schemas that ones uses to describe schemas.

Jordan Brown (talk) 19:06, 3 April 2020 (UTC)[reply]

This dab page does list database schema and XML schema. YOu are welcome to write a generic article., even a short stub, if you have reliable sources to cite. Staszek Lem (talk) 20:20, 3 April 2020 (UTC)[reply]
Yes, I saw the database schema and XML schema links. But I'd say that they are two specific instances of a more general concept, and it is that concept that should be linked from here (and various particular schema styles would be linked from there). And yes, I know how to create an article, but right now I don't have the energy to do the research. Jordan Brown (talk) 01:07, 4 April 2020 (UTC)[reply]
From what I understand after a bit of reading, the word "schema" is just an obfuscation of the term "model", in the meaning "a model of the representation of data in a particular application domain from a particular point of view". Or, alternatively, a botched attempt to distance from the generic-sounding term "data model". As a result, in each case of "X schema" the specific definitions have nothing in common beyond the generic philosophical meaning of the term. And as usual, the wikipedia article data model sucks. And by the way, "metaschema", at least as defined in a couple places I googled is nothing but an "internal schema". Whatever. As one of the 2004 reports on database standards says there is no standards for schemas. May be things changed as of 2020. Staszek Lem (talk) 05:22, 6 April 2020 (UTC)[reply]