[Sigia-l] multiple search index usability/conceptual model

[Ziya Oz]

"Eric Scheid" wrote:

>> Technically it'd be quite straightforward, but tedious
> 
> It would only be straight-forward if terms were unambiguous in any
> overlap between contexts.

As evidenced by the word 'technically' above, I'm referring to the
architecture of the search UI not the contents of the query. The reason why
SQL query aggregation/abstraction works across multiple DBs is because
there's a semblance of standards--optimization and variations being taken
care of behind the scenes by the front-end interface.

One can also use 'redirection' (much as outlined in the MSFT taxonomy tool
recently mentioned here) where variations in different domains/context are
mapped to common/global entities. As I said it's a matter of another layer
of abstraction.

If there isn't significant commonality among different DBs, maybe presenting
each DB in its own UI is the way to go. The UI look and feel as well the
supporting widgets may be standardized so that the user has to deal with
(hopefully) a smaller number of variations among UIs.

Again, intellectually, this presents virtually no challenge; it's just a lot
of pedestrian work trying to deal with exceptions, rules, variations, etc.
It's like painting a wall with a toothbrush.

Best,

Ziya