[OAI-implementers] usefulness of free text fields

Glen Newton gnewton@orca.cisti.nrc.ca
Thu, 26 Jul 2001 16:52:05 -0400


Hi,

I have a question, which will perhaps reveal my nascent knowledge and
in this area:

OK. Let's look at an example <usage> metadata (from the 1.1 spec):

Example 1:
<usage>
    Metadata may be used without restrictions as long as the OAI
    identifier remains attached to it.
</usage> 

Yes, I can understand the content of this. But when my harvester goes
out and gets 10 million OAI metadata records, I am not going to go
through all of them by hand to figure out what their usage
is. However, if I want to be able to legally present my harvested
metadata, I DO want to be able to differentiate different usages
(public domain, resticted to licensed users, etc...). But the present
OAI does not seem to let me (easily automatically) do this (unless
this can be buried in a DC qualifier...). 

What would be useful would be a numerical attribute for usage, which
would act a a look-up for a list of values maintained by someone. For
example: 

Example 2:
<usage code="1">
    Metadata may be used without restrictions as long as the OAI
    identifier remains attached to it.
</usage> 

This code value would be absolute, i.e. I could write applications
which could use the code value to drive it. Also, although I can read
the above usage content (example 1), if it were written in Russian or
Maori i would not be able to understand it. 

The table could look like this: 

Code   Meaning
0      Public domain
1      Metadata may be used without restrictions as long as the OAI identifier remains attached to it.
2      Available only to Cornell staff and students
3      Available only to licensed users of Elsevier journals
4      Available only to licensed users of the Elsevier journal "Computers & Geosciences"


Now this "solution" is rife with many problems: who maintains the
list; the codes will not be mutually exclusive, but this may not be
obvious; and it is starting to look a little like SFX. 

I may be missing something important here, so please illuminate me...

thanks,

	glen