Communicated from Nan Gelhard: 12/13/2003 4:33 PM
Comments on in ORB as a
knowledge management tool:
1) Human beings use language built from context and
experience. Language is not static. I read old books, and so I use old,
contemporary dictionaries to translate the language. I group books by subject
matter, authors' point of view, authors' geographic location, and published
dates in order to build sense.
2) Humans understand categories and contextual meaning.
3) Computers are just swift idiots.
4) Humans jump to conclusions based on their experience and
the context of the conversation. When we talk we negotiate the context. (For
instance, "Did you see Seinfeld last night?..." )
The InORB utility gives authors
the ability to specify meaning to strangers (researchers, customers) by simply
specifying 250 or so words or phrases of a controlled language, the author's
voice. A motivated author can define thesauri for those words or phrases and
make his work accessible to neophytes as well as experts.
An initiate knows the local,
technical or fashionable language, a neophyte knows a more general term. For
instance, in my industry a helmet is sometimes called a skidlid, vehicles are
shod with tires, and a supercharger is sometimes called a blower.
Three words can specify a
meaningful context. Ravens, Poetry, Poe has a specific meaning that is
different from Ravens, Cleveland, Baltimore. Which is different from Ravens,
Corvids, scavengers.
A user can specify his three
words in his own context (from his experience and understanding) and get
meaningful answers.
This is what is missing from
current database search tools. There is no simple mechanism for the basic
mechanics of communication, negotiating the subject. A tool like InORB lets
authors and users with no particular knowledge base training specify context.
Nan Gelhard