forum SignWriting List Forum
  Message 6576  |  Previous | Next  [ Up Thread ] Message Index
From:  Bill Reese
Date:  Thu Mar 21, 2002  2:56 am
Subject:  Re: SignBank 3: What kind of search-routines do you want?


I know that when I do searches on the net, if I want two words to be together, I
enclose them with quotes. For example, if I wanted to search for "Pink
Grapefruit" available in "Florida", I could write my search as:
"Pink Grapefruit" Florida
The "Florida" wouldn't need to be with "Pink Grapefruit" but "Pink" and
"Grapefruit" would.

I don't know exactly how this would apply to searching the sign bank, but the
nearest application I can think of is if you have two symbols that you want to
together, perhaps in a sequence - one right after the other - and a third
that could be anywhere.

Of course, I would also think that the results would be best displayed by which
signs satisfy as many of the search symbols as possible. But I'm sure Todd is
probably doing that already.


Valerie Sutton wrote:

> SignWriting List
> March 20, 2002
> Hello Stuart and Everyone!
> I am printing your suggestions for search-routines for Todd, our
> programmer, and we will see what we can do for SignBank 2002. I am
> just happy that we can search by symbols at all......SignBank 3 was
> harder to program than most people would realize....
> Thanks very much, Stuart, for your input. I hope others will tell me
> what kind of search-routines you need too...Val ;-)
> ----------------------
> Stuart wrote:
> >Sort of ... I guess maybe my idea is more "convoluted" than the other search
> >routines. For example, in ASL, the sign GO, is 2 index hands moving in the
> >direction of the action. Now, there are "inflections" that change that
> >sign, such as repetitive movements which give the idea of going frequently,
> >or once in a while, or regularly, or (with appropriate facial expressions)
> >going tiredly or whatever. My idea for the "morphology module" concept is
> >if there are standard movements or facial expressions which tend to modify
> >signs to produce a standard change in meaning, maybe those could be
> >catalogued in the morphology module and then when I put in the symbols for
> >go, the morphology module could also look for signs that employ these
> >"modifications" to give me alternative meanings or related signs. Like "to
> >attend" is usually the sign for "to go repeatedly". That kind of thing.
> >
> >Maybe it's something for SignBank 2004 <grin>
> >
> >Stuart
> >
> >Stuart Thiessen
> >Des Moines, IA
> >----- Original Message -----
> >From: "Valerie Sutton"
> >To:
> >Sent: Wednesday, 20 March 2002 11:52
> >Subject: Re: SignBank 3: What kind of search-routines do you want?
> >
> >
> >> SignWriting List
> >> March 20, 2002
> >>
> >> Stuart Thiessen wrote:
> >> >Another possibility is to do what we might call dictionary vs.
> >"inflected"
> >> >forms. For example, if I put in the sign "TO GO", could it find "I GO",
> >"HE
> >> >GOES", "THEY GO". Or repetitive motion signs or noun variants of verb
> >> >signs. This starts to get out of the phonetic aspects and gets into
> >> >morphology of signs, but maybe there is a way to add a "morphology"
> >module
> >> >to the signbank so that this kind of research would be possible?
> >>
> >>
> >> Thanks for this input, Stuart! Wow....a "morphology module"...that
> >> sounds sophisticated!
> >>
> >> Regarding what can be done right now in SignBank....In SignBank 1,
> >> Searching by Words, you can type in a search-word, like "GO", and if
> >> there are 30 signs that are variations of "GO" in the dictionary,
> >> SignBank 1 finds them all and lists them with words. Then you click
> >> on the word-variation you want, and you get that sign... Take a look
> >> at this example, with the sign for "HELP" listed all the HELP
> >> variations. Is this what you mean for the "morphology" module?...

  Message 6576  |  Previous | Next  [ Up Thread ] Message Index