Re: word tools

From: Eric Lease Morgan <emorgan_at_nyob>
Date: Fri, 4 Apr 2008 12:08:12 -0400
To: NGC4LIB_at_LISTSERV.ND.EDU
For a good time, I think I'm going to create something tentatively
called OPAC Helper, and it will be a word tool. The purpose of the
tool will be to aid people in the use of library authority records.
This is how it will work:

   1. Enter a word or phrase to search or browse a
      dictionary/thesaurus of authority headings.

   2. Search/browse results will return a number
      of characteristics such as:

       o heading with record and class number
       o definitions
       o synonyms
       o see from references
       o see also references

   3. Go to Step #1 until a relevant heading
      is identified.

   4. Search and display records from a remote
      catalog by combining the authority
      headings with standard query APIs
      (such as SRU).

   5. Go to Step #1 until satisfied.


To implement this idea I will:

   1. Download authority data from FRED. [1]

   2. Download dictionary/thesaurus data
      from DICT. [2]

   3. Read, parse, and amalgamate the authority
      data with the dictionary/thesaurus data
      into locally defined XML snippets (or
      maybe TEI can be used).

   4. Index and make accessible the XML using
      the DICT protocol. [3]

   5. Write a Web-based program that implements
      the user interface, above.

Because the authority records will be accessible through the
"standardized" DICT protocol, and because searches against the
catalog will be through standardized query interfaces, the whole
thing should be implementable in a very modular fashion. The
prototype will be just that -- a prototype. It will not be intended
to be a "next generation" library catalog, per se. Instead it will be
intended to demonstrate: 1) one way to use the "cosmos" created from
the systematic classification process, and 2) how interfaces can be
cobbled together using Web Services computing techniques. If the
prototype is successful, then the output of the DICT server could be
integrated into other applications.

It is not all about cataloging, and it is not all about information
retrieval. Instead, it is about combining the best practices of both.

If anybody wants to help, then just let me know. On our mark. Get
set. Go.

[1] http://www.ibiblio.org/fred2.0/authorities/
[2] http://www.dict.org/
[3] ftp://ftp.isi.edu/in-notes/rfc2229.txt

--
Eric Lease Morgan
University Libraries of Notre Dame
Received on Fri Apr 04 2008 - 10:44:29 EDT