From lpsmail@access.digex.netThu Nov  9 15:48:57 1995
Date: Thu, 9 Nov 1995 11:11:54 CST
From: Shipment <lpsmail@access.digex.net>
Reply to: Discussion of Government Document Issues <GOVDOC-L@PSUVM.BITNET>
To: Multiple recipients of list GOVDOC-L <GOVDOC-L@PSUVM>
Subject: ADNOTES: COUNCIL: SCOUT FOR GOVT INFO IN THE WEB

----------------------------Original message----------------------------
      THE MESSAGE BELOW IS FROM ADMINISTRATIVE NOTES,
      VOL. 16, #15  (Nov. 15, 1995).


      SCOUT FOR GOVERNMENT INFORMATION ON THE WEB

      Remarks by Raeann Dossett
      Internet Specialist, Electronic Transition Staff
      at the
      Fall 1995 Depository Library Council Meeting

          Scout is the working name for a group of
      Internet tools that will present a web-based
      interface for searching Federal government
      information on the Internet.  We aren't using this
      designation because we want to introduce jargon into
      the project, but as a way to keep this aspect of the
      larger Pathway Services project distinct from the
      other components Maggie delineated for you.  I hope
      the term Scout is useful in that regard this
      morning.

          If you are familiar with web tools such as
      Lycos or InfoSeek, you already have a good idea what
      Scout is going to do.  Those tools allow you to
      search the Internet, especially the World Wide Web,
      by using keywords and various operators, such as
      Boolean operators.  As a piece of software, Scout
      will look similar to these existing tools.

          Behind those looks, however, Scout will deal
      with quite a different set of information.  Instead
      of trying to broadly index the entire Internet, we
      will focus Scout specifically on government
      information on the Internet.  In general, this will
      mean restricting its activities to sites in the .gov
      and .mil domains only (e.g., www.access.gpo.gov,
      www.navy.mil).  I say  in general  because we will
      make exceptions for official government sites housed
      at educational (.edu) or other sites outside the
      .gov and .mil domains.

          Here is a sketch of the software pieces that
      will comprise Scout:

               *    A web crawler OR a broker/gatherer:
      This is the tool that will go out on the world wide
      web at our direction, obtain information about the
      content of Internet sites, and bring that
      information back home.  That content will be:

                    --   documents or files
                              ASCII     SGML      GILS
                              HTML PDF

                    --   directory level information for
      groups of documents or files stored in various
      Internet protocols
                              FTP  gopher
                              WAIS HTTP

               *    That information will be filtered
      through another piece of software that will discern
      the filetypes and create a fielded database.

                    -- fields such as:
                              Titles    URLs Keywords
                              Originating  agency

               *    A database search engine.  It will
      perform boolean searches, as well as allow for
      natural language queries.  You will be able to limit
      your search to a particular field, such as title,
      originating agency, and URL.

               *    Finally, because Scout is a web-based
      tool, Scout will provide an active link to the
      information resource--click on the URL and you go
      there.

          Beyond restricting the scope of Scout to
      government information, and going into those domains
      as deeply as possible in search of actual content,
      we will be able to manually enhance the records in
      this database to make it more useful to depository
      libraries.  We will be able to add additional
      fields, such as a title tracing field for products
      that have historically been supplied in paper
      format.  We will also be able to augment existing
      fields--providing keywords in addition to those
      created automatically, for instance.

          Currently, we have a prototype Scout running at
      GPO, and we re evaluating additional software
      packages.  Since Maggie and I are on a tight
      deadline--we only have 10 months left at GPO--we re
      proceeding on this project as fast as possible.  We
      hope to be putting the system through a beta-test
      within the next six weeks.  You all will be able to
      see it as soon as we complete the requisition
      process for the software, and work out all the major
      functional kinks.

          I hope this gives Council more information
      about this aspect of the Pathway Project.  We would
      welcome comments from Council, and from members of
      the audience, on both our plans for the system and
      its usefulness as it evolves over the next few
      months.