RE: Content Classifications / Taxonomies

Again, in my last job, we wrote a content management system from 
scratch in .NET that grabbed its content from Word documents, 
which the theory was, the documents could then be taken from 
seperate 3rd party document management systems. We made a bit of a 
blunder by storing the documents inside sql server however - i 
definitely do not recommend storing binary inside a database!

The related links ala amazon, and news sites, does seem to be 
quite a major benefit to being able to define custom metadata 
however, more so than the searching I'd say.

On Wed Jun 15 08:46:27 PDT 2005, "Adriaan M. Bloem" 
<[email protected]> wrote:

> Chris,
> 
> That's an interesting question. I was wondering the same thing.
> 
> I'm currently implementing a CMS at our university and we want to 
> be able to
> categorise the content to achieve mainly two goals: first, to 
> make it not
> only searchable (in the old-fashionable full-text way) but 
> actually findable
> (if there is such a word); and secondly, we'd love to do some 
> fancy
> automatic references with pages - somewhat in the fashion of 
> Amazon, which
> has a "people that bought this, also bought..." - so pages which 
> aren't
> quite what people want still have links pointing them in the 
> right direction
> (i.e., pages with comparable content).
> 
> Now of course this means categorising content, and this has to be 
> done a lot
> more structured than plain asking users to fill in some keywords 
> with the
> content they're adding. Which means we have to build at least a 
> small tree
> of classifications.
> 
> On the other hand, I've been involved in some projects trying to 
> create a
> full ontology of, for example, legal terminology (LEXML). Which 
> of course is
> an impossible feat to accomplish. Even cutting it down 
> pragmatically for
> knowledge management purposes can be labour-intensive, cumbersome 
> and
> eventually so complicated previous projects trying to do this 
> never amounted
> to a functional implementation. As you can understand, I'm weary 
> of going
> down that lane again.
> 
> So what I'm looking for now is just a pinch of taxonomy and a 
> whiff of
> classification, but for pure web content management purposes I 
> don't think
> it'll be worth it to go any further than that. And I think you're 
> right in
> saying that trying to go any further probably amounts to using a 
> CMS to do
> something it's not. But the basics should be there - you don't 
> want the
> current chaos that makes up the university website (between 
> 100,000 and
> 200,000 pages of content on-line without any real structure).
> 
> 
> Adriaan M. Bloem
> 
> __
> 
> Project Leader Implementation CMS
> ICS Webcommunication
> 
> Internet Coordinator
> Faculty of Law
> 
> Leiden University
> Visiting address: Steenschuur 25, room C013
> Postal address: P.O. Box 9520, 2300 RA Leiden, The Netherlands
> Telephone: +31 71 527 7897
> 
> 
> 
>> -----Oorspronkelijk bericht-----
>> Van: Chris
>> Verzonden: woensdag 15 juni 2005 16:21
>> Onderwerp: Re: [CMS] Content Classifications / Taxonomies
>> 
>> In my last job I did some work for a british education 
>> institution who were very keen to organise their content in the 
>> 3rd party java CMS system we were tasked with implemented. They 
>> loved their metadata and categorisations.
>> 
>> What interested me though is at what stage CM, content 
>> management, verged into what is basically systems for document 
>> management or perhaps even software specifically for libraries. 
>> I understand CMS in an old fashioned way in that it allows 
>> people to publish websites without technical knowledge (HTML 
>> etc.). So I guess my question is, do people expect metadata 
>> categorisation in modern CMS systems or is this just been bolted 
>> on to make CMS systems do yet another job that is the realm of a 
>> different software market?
>> 
>> chris
>> 
> 
> 
> 
>