- From: MacKenzie Smith <kenzie@MIT.EDU>
- Date: Fri, 09 May 2003 07:03:28 -0400
- To: www-rdf-dspace@w3.org
At 01:27 PM 5/8/2003 -0700, Bass, Mick wrote: >- How do users interact with the vocabulary (i.e. how do they select their >desired term(s) from the N00,000+ that are in the vocabulary)? In >particular UI/selection/suggestion/search techniques. I've never done this personally, but my understanding is that it's usually done "asynchronously" with the data entry or cataloging for the item in hand -- i.e. you have a copy of the thesaurus in print or CD or on the Web with its very own UI and you consult it when you need a term. Most catalogers are intimately familiar with these vocabularies and often with the domain they come from so it isn't considered a huge burden for them. >- How are the vocabularies maintained? i.e. with 100,000 entries what >processes are used to identify which terms need to be >reviewed/updated/merged/deprecated etc. and what metadata do such >processes generate? If SIMILE is to offer facilities for vocabulary >maintenance, what function points are likely to be required? Again, hearsay that I can verify if necessary. By teams of experts. Manually. Very, very slowly. With lots of process and QC and trust systems for who's allowed to add/remove/change entries. They often have official editors and reviewer, kind of like a dictionary or encyclopedia. >- How are the vocabularies today shared across systems (i.e. what transfer >formats are used)? For the big, official ones (e.g. MESH and LCSH) they can be acquired in digital formats for local loading but I have no idea if they're encoded in XML or what... I can find out if you need to know that. MacKenzie/
Received on Friday, 9 May 2003 07:07:35 UTC