User talk:Genezistan

From Glottopedia
Revision as of 08:53, 23 November 2008 by Genezistan (talk | contribs) (New page: ==Change in Presentation== Hi, Because Glottopedia has just been started, I wonder if you could consider my proposal as far as its wiki structure is concerned and as it is explained belo...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Change in Presentation

Hi,

Because Glottopedia has just been started, I wonder if you could consider my proposal as far as its wiki structure is concerned and as it is explained below:

A Wiki format is typical of the organisation of lexical knowledge on a morphological basis (alpha sort) supported by hypertext links of various sorts to move in the text following another sort. Now the categories used in any collection of wikies tend to follow a hierarchical fashion despite the fact that human language is not composed of elements that can be arranged/sorted into hierarchies, rather, they are best described as a set of recursive elements that are easily described by using a BNF notation.

Now besides lexical knowledge we have procedural knowledge, which is in fact the "how to" and the know-how type of representation of knowledge. As such, it is regarded more valueable and useful than lexical knowledge which is usually acquired through rote and incidental learning as opposed to procedural knowledge, which become competences through practice and the use of non formal logic. Formal logic has taken semantics away from linguistics and has grossly simplified its mission to help design robots and recently WordNet (absolutely hopeless), and still word hierarchies that are a mistake in thinking.

For instance in a recent article in Cognitive Systems Research [Economically organized hierarchies in WordNet and the Oxford English Dictionary]the author reveals that he is not aware of the concept of definitions and therefore he still believes that he is dealing with hierarchies.

Now core or upper ontologies in AI also reveal that the current model of semantics of a human language (the most important semantic primitives) are not properly understood or defined. This is why they cannot integrate or harmonize ontology languages.

I claim that by using objects, properties and relations only, you can solve the problem. Of course, you need to have time, space and motion, or cause and effect also defined, but that can wait. The point is that relation should be used to change the lexical character of a wiki into a more transparent format where relations are also exposed. To remind you, relations are basically verbs, they represent the ultimate connection between noun phrases or names, etc. (headings, titles and labels) and offer you a lot richer world than the Boole algebra.

If you do not see the point, look at the example of a website of [Elsevier] where the selection of views offers you an example of the relations that I am talking about.

I am willing to expand on the subject, should you be polite enough to ask me to.

Frank 09:45, 23 November 2008 (CET)