RE: FC Representation/Deliverables

>  > Here's a question about the utility: does it let you build 
>  one tree, 
>  > or multiple trees (ie. facets)? It sounds pretty interesting! I'm 
>  > gonna check it out.
>  
>  Peter:
>  
>  OK, heres where my ignorance gets in the way. If multiple 
>  trees = multiple facets, then the answer is kind of iffy.

Each facet is indeed an independent tree (hierarchy).

>  This utility is primarily of an "encoding" type. You can 
>  create any number of hierarchies, any number of levels deep. 
>  You can have the same word appear more than once, though the 
>  utility is "cognizant" of this and will prompt you that you 
>  are creating a second instance of the word.

That's cool. 

>  When you search for this word later within the hierarchy, it 
>  will bring up all instances and show you the lineage of that 
>  branch of the tree and you simply click on the one that you 
>  are after.

Cool!

>  Again, as this tool is primarily for encoding the data into 
>  the IPTC meta data which is then embedded in the header 
>  portion of the image file, you can always choose multiple 
>  branches to add to an image. In that sense it allows 
>  multiple "facets" to be added.
>  
>  For instance. If I have a photograph of a cow in a pasture, 
>  the first two branches that I would pick would have to do 
>  with these two descriptors. If cow falls in the subset of 
>  "livestock" which is, in turn a subset of "animal" then all 
>  three of those keywords are added to the IPTC info. If 
>  "pasture" is a subset of "farm", which is a subset of 
>  "rural", which is a subset of "scenery."

So only the keywords are shared, in a list like "animal, livestock, cow,
scenery, rural, pasture". Is that it? In that case, the receiving end
can't use the taxonomies for browsing unless those are shared as well
(maybe using XFML).


>  If you instead consider that a particular set of branches = 
>  one tree, and want to build a completely different tree, you 
>  can "backup" or "export" that set and either load a new one, 
>  or create a new one from scratch.

I didn't get this last bit :(

>  Whether or not it's a useful tool to others building faceted 
>  classifications, I'm not really sure.... but I'm sure you 
>  would know simply by giving it a twirl.

I'll try to find some time today..
Peter



----------
Thanks for playing. To unsubscribe from this group, send an email to:
[email protected]

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/