[Date Prev] [Date Next] [Thread Prev] [Thread Next] Indexes: Main | Date | Thread | Author

Re: [ba-ohs-talk] Keyword Indexing to Improve Email and IT


On Monday, April 22, 2002, at 12:19 AM, N. Carroll wrote:
> John,
>
>> Getting people to add keywords is a huge stretch.
>
> That depends on one's command of interface and knowledge
> of user psychology as much as it does the nature of the users.    (01)

It seems to me that the best approach to adding metadata to e-mail or to 
other information items is to provide highly handy tools that can be 
nudged by people. Ie, bits of AI type routines and agents, but more 
importantly the ability to really quickly, easily, and effortlessly work 
with these, guiding them. Maybe this is obvious: symbiotic rather than 
artificial intelligence.    (02)

So as you say, "much depends on one's command of interface".    (03)

Actually, a very simple example of this is the process of creating an 
e-mail header for a message: 100% user-driven. But when someone replies, 
the RE:<original message header> header is automatically composed (if 
you're into PR, you can call that routine an agent :-). Then, simple 
mail clients can sort by subject and organize threads, driven by users 
clicking on column headers.    (04)

I'm assuming, by the way, that unrev e-mails are redistributed to the 
subscription list, which include's Eugene's program to stamp purple 
numbers on them. If they hit Eugene's program first, and _then_ got 
redistributed, we could gradually start receiving richer versions of the 
e-mail items in our own mailboxes, For starters, arriving with the URI 
of message in the archive. I could then easily grab and feed that to my 
personal information husbandry routines.    (05)

Cheers,
Mark    (06)