Cet article examine quatre sites d'accès
Internet fournis par le gouvernement du Canada et certains outils
de recherche disponibles à ces sites : a) le site Internet
principal du gouvernement du Canada, b) le Centre d'information
en direct intergouvernemental (un projet intergouvernemental),
c) le site WWW de la Bibliothèque nationale du Canada,
et d) le Gouvernement Ouvert. Les outils de recherche et de localisation
mentionnés incluent GIFT (la technologie de recherche d'information
du gouvernement), Champlain, et Hudson.
In 1993, when I first started to incorporate government
information on the Internet into my work as a reference librarian, the
hardest part was finding it. Many hours of Gopher "tunnelling" uncovered
a few nuggets of information buried in sites more often than not "under
construction". There was no "official" government site, you could not
tell from the Internet address whether a site was a government one, and
some government departments piggybacked on commercial servers. There
were no comprehensive collections or gateways to Canadian government sites
on the Internet, and the best collection of Canadian government documents
in full text was at an American Gopher site called wiretap.spies.com.
Much has changed since then. The quantity and quality of the
government information available has increased enormously, although
unevenly. The World Wide Web is now the medium of choice, and several
government initiatives designed to make it easier to locate government
information on the Internet have resulted in significant improvements.
For example, there is now an "official" government site, federal and
provincial government URLs are standardized and recognizable, the
Weekly Checklist of Canadian Government Publications
3 includes
the URL for documents that are available on the Internet, search and
retrieval tools specifically for Canadian government information have been
developed, many government sites have organized searchable documents
collections, and there is a single-window, gateway site to
information from all levels of government on the Internet.
In this article I will describe the four sites where most of these
developments are unfolding:
This, the "official" Government of Canada site, was launched December 14,
1995.
In the news release announcing its official opening, the "Canada Site" is
described as providing Internet users "in Canada and around the world with a
single electronic access point to general information about Canada and to a
wide variety of federal government information and services."
4
The general information about Canada is covered by fact sheets on
Canada and the provinces and territories, a map of Canada, pictures and
brief descriptions of Canada's symbols, by a text describing how
government operates, and by a brief overview of, and links to, some of the
key institutions of the Government of Canada. The section on Programs
lists a few federal programs and services, but many more links to sites
with government programs and services are accessible through the Services
Catalogue at the Intergovernmental site mentioned below.
The Primary Site (as it is also known) works well as a gateway to
government information at all levels, not just federal. "Federal
Institutions" is an alphabetical list of federal departments, agencies,
commissions, Crown corporations, etc. Where a public Internet site exists
a link is provided, otherwise the link is to a brief description of the
institution. Under "Other Governments" is a link to the official
provincial government sites and to the Intergovernmental On-Line
Information Kiosk, which provides access to municipal sites as well. As
explained in the December 14 news release, the intention is: "that all
existing government information and services can be accessed through a
single Internet address."
Besides this gateway function, one of the most useful features at this
site, is the "What's New" area, which covers not just what new information
has been added to this specific site, but also the new services,
publications and news releases added to WWW sites across the federal
government. It is not clear how often or how comprehensive these
"regular" updates are, but it is a great advantage to have new "releases"
listed in one spot. Another good idea, not often seen on the Internet, is
that the "What's New" area is archived. Currently the Archive holdings
begin with December 1995, and are updated monthly. I hope that the
intention is to retain all of the information in the Archives. Beyond its
updating value, it provides an interesting overview of activity on the
Internet over time, which will always be of interest.
The documents located at this site are keyword searchable. It is
commendable that searching instructions are given, although the
definitions and explanations of Boolean operators are the most confused I
have come across. The terms used to describe AND and OR -- "intersection"
and "union" -- do not help much, and an error in the nested query example
doesn't help either. The documents retrieved are first displayed as a
list of titles, with the size and type of file given. Unfortunately, when
the full text is displayed, many of the files show no author, date,
department or section responsible for the text. This is a common problem
on the Internet (not specific to this site), but one that could be
remedied if every webmaster required this information to be present before
agreeing to post a text. Information retrieved in this way (without
source information, date, etc.) is almost useless for academic research
purposes.
Besides being able to search this site by keyword, GIFT (Government
Information Finder Technology), the new search and retrieval tool for all
Canadian government information on the Internet, is located here in its
initial phase. (There is also a link to Champlain, a keyword searching
tool, described below under the Open Government Project. These two search
and retrieval systems work differently and will retrieve different texts.)
This government network, currently under development by Government
Telecommunications and Informatics Services (GTIS) of Public Works and
Government Services Canada, is intended to help you locate all authored
government documents from departments, committees, agencies, etc., with
the results supplied in a readable and easy to understand format. From
the Primary Site you will see listed several keyword-searchable government
sites and documents collections, which can be searched individually. More
sites are being added on a continuous basis.
At the GIFT home page there is a similar list of individual collections of
government documents or sites organized under broad subject categories. You
can search the sites separately, or the combined sites under a subject.
Clicking on GIFT, for example, will allow you to search the documents posted
at the GIFT site. Clicking on its subject heading "Internet" will allow you
to search GIFT and the other sites under that category as a combined set.
Click on "Search Terms" for instructions on how to construct a search query
(e.g. phrase searching requires quotation marks).
But the real power of GIFT is only evident when you select "Search
Other Sites", and then select "Z39.50 Compliant Multi-site, Multi-database
Searching." 5 This is where the next four phases
of GIFT are being developed. Although still "under construction", it is
more or less up and running, and undergoing testing and evaluation by GTIS
and its working committees. The official release date is scheduled for
the Spring of 1997. In a Focus Magazine article about GIFT
posted at the CRTC WWW page, GIFT is described as "transparent and
seamless"; "users can request GIFT to find instances of a certain subject
throughout the whole domain of electronic government documents without
mapping their way through department databases or manipulating the results
to make them intelligible." 6
This user-friendly , "single window" concept, bringing together all
electronic
texts from the hundreds of existing government locations, to be covered by a
single search, sounds like a dream come true. If all goverment bodies
contribute (this is a voluntary initiative) it could well be. All indications
are that it will succeed. The manual work involved for the information
providers is minimal, although the documents remain at their original site,
where the responsibility to maintain and update the information rests.
It works something like this: A short record of basic elements
describing each document (containing information like the author, date,
source, language, keywords, subject terms, etc.) is created, and this
"document profile" is what the GIFT server collects and searches. A web
crawler collects these profiles at regular intervals. Based on the Z39.50
protocol, the GIFT server can then retrieve the full text documents from
many other kinds of servers (WWW, Gopher, FTP, Fulcrum Surfboard, WAIS,
etc.), translate from a vast array of file formats (image and text), and
display the documents in the format chosen by the user. There may also be
pointers provided to related non-electronic documents.
This certainly sounds very promising. It seems to me, however, that
despite the impressive technology, the quality of the search results will
only be as good as the document profiles allow. In a document describing
how GIFT works, called "How to be Transparent and Seamless", the author,
Oliver Javanpour of GTIS, Project Manager for GIFT, explains: "the
document profile is generated automatically by connectivity tools that are
developed by the GIFT Project," and later in the same document he says
that "discriminating terms" ("unique" keywords) are automatically
extracted. He also says that: "The entire premise of this mechanism is to
relieve the author and informatics personnel from the burden of producing
complex document profile information." 7 All this
suggests that the whole process would be automated.
"Controlled terms", one of the twelve suggested fields for the document
profile listed in the same document caught my eye. There is not much
detail on how these will be generated. According to Mr. Javanpour, there
is a commitment to including some form of subject indexing or controlled
vocabulary, even though it will require some measure of human (librarian)
intervention. It has not been decided yet whether they will try to create
a built-in, automated thesaurus to generate these, or if librarians will
manually assign these terms.
As GIFT appears now, you can see the list of sites or documents
collections
available so far, each of which you can choose to include or omit from your
search by clicking on the box in front of each database. If the sort mode is
set to "by database" (the default), the display will show a list of the
databases retrieved, by name, followed by the number of documents found in
each database. If you choose a site that has several documents, a list of the
document titles or WWW page names will be displayed with no other information
to help in identifying the source or the content. A bit more information at
this point (as is provided in Champlain's display) would make GIFT more
efficient to use. Unfortunately, like many other automated search and
retrieval tools on the Internet, there is also quite a bit of duplication at
this point, which will become more of a problem as more databases are added.
This site is described, on the "About this Service" page, as "a
single-window on Federal, Provincial, Territorial and Municipal government
on-line services in Canada." It is maintained by the Canadian Governments
On-Line (CGOL) intergovernmental project team on behalf of all levels of
government, and is meant to "provide the public with an easy access point
to all information and services provided by governments across Canada." 8 Although this sounds much like the Primary Government
Site, it is different in several ways.
The home page provides a menu with links to each government level in
Canada
and to some foreign governments. Each sub-page for the provinces then has the
official site for that jurisdiction, the home page of the head of the
government, documents related to the information highway, a list with links to
the institutions of higher education in that jurisdiction, weather forecasts
for the area, contacts in the government, and phone, fax and e-mail
directories, if available. The municipal sites are listed in alphabetical
order, and by province, with direct links to the sites. (At this point, not
all of the municipal sites listed are necessarily government sites, or even
have any government information.)
The documents actually posted in the "Documents Library" are intended
to be "generally restricted to content of an intergovernmental interest or
public service," or "documents or reports that have an impact on
government on-line services in general," and not duplicating information
provided by any other official Canadian government source.
9 Currently these are documents related to the information highway,
and reports and other texts relevant to the CGOL initiative itself. These
documents can be found using the index, by document title, or by
jurisdiction. (A welcome touch is that for each document listed, the size
of the file, its language and format are shown in brackets after the
title.)
The Services Catalogue is a list of on-line services offered by all
levels of government across Canada, which also briefly describes and
classifies each "service". (Services seem to include any public presence,
whether a WWW site, BBS, 1-800 number, on-line kiosk, database, etc.) One
of the main purposes of this service is to facilitate information sharing
between all government institutions across Canada at all levels, to help
avoid duplication in setting up and providing services on the Internet,
and to help determine the "best practices" to follow.
The Catalogue is searchable by jurisdiction, type of service, subject,
or government department. The "subject" categories are very broad (e.g.
"Finance" is the subject term for the Department of Finance WWW site).
This may be adequate for the above-stated purpose, while there are not too
many entries, but it makes the Catalogue less valuable as a tool for the
public trying to find out which on-line service to access for a specific
need, especially, in the future, when there will likely be many more
entries than there are now.
The template used to describe the services in this catalogue is similar
to the
"Document Profile" being developed by GIFT. The "Overview" section is
different though, and could be one of the Catalogue's most useful features.
Very few entries have this section filled in yet, but the ones that do, show a
very brief description of the contents and purpose of the database. The
"Access address/path(s)" sections have not been filled out yet, but in the
case of Internet resources, the intention is that the URLs will be given as
hypertext links allowing for a direct connection. The "Access Pricing" and
"Methods of Payment" sections are thankfully also blank at this point. The
Catalogue is being compiled from the responses to a survey sent out to
government institutions in 1995. It has not been decided yet how this
catalogue will be maintained or updated, beyond the feedback obtained from
government service providers via on-line forms at the site. According to
Claude Bourgeois, InterGov Director, Provinces and Territories section,
decisions on how to proceed will be made after the DMR Group presents its
final report on the project.
The National Library of Canada is in the process of migrating
information from
their Gopher to their WWW site. Some of the information, such as the list of
government sites, will be dropped in the process, as this is now being done
more thoroughly by the Primary and Intergov sites. One of the initiatives
they are working on now, which will assist in locating government information
on the Internet, is the "Canadian Information by Subject" service. The
objective of this service is to provide links to information about Canada,
arranged by subject, from Internet resources, wherever on the Internet they
may be. This includes, but is not limited to government information.
The subject arrangement is based on the Dewey Decimal Classification
System,
but it is not necessary to know the system to be able to benefit from the
arrangement. The Subject Tree can be browsed in subject order or in
alphabetical order. What is "classified" is an entire WWW site or
database of
resources, not individual files on the Internet. For example, under the
subject heading "Genealogy" are links to National and Provincial Archives,
Genealogical Societies and Associations, and to guides and resources posted by
individuals. The reliable subject arrangement and broad scope make it a
useful and popular tool. The fact that government sources are combined with
others in this subject "index" makes the government sources more accessible
for those users who may not have thought of the government as a likely source
for the information they are seeking. The subject classification is also much
more useful for researchers unfamiliar with government structure, and just as
important is the added browsability that a subject tree affords. Work on this
service is ongoing. More sites are added to it on a weekly basis.
Other initiatives in the works at the National Library also include
government
documents, according to Nancy Brodie, of Information Resource Management, at
the NLC. For example, a new government section has been added to the
Library's WWW site, intended specifically for its primary clientele (Canadian
libraries). This collection has links to key federal and provincial
sites, indexes, and documents, such as government publications catalogues.
The Electronic Publication Pilot Project, set up to examine the issues and
challenges involved in cataloguing electronic publications, also included some
government serials, and now, in the National Library's electronic documents
collection (the continuation of the EPPP), government documents are among the
electronic publications acquired and catalogued. These publications can be
located using the AMICUS database.
Developed as a pilot project to "provide greater access to government
through
information networks," 10
OpenGov was
instrumental in getting government information from and about institutions
such as the House of Commons, the Senate, and the Supreme Court on the WWW.
Although they now have their own "home" on the Parliamentary site, OpenGov
still maintains a link to these and other government resources.
As a skunk-works site for federal government initiatives on the Internet,
OpenGov is also continuing to develop new services and service-delivery
products. Champlain was announced in April 1995. Their next project, still
being developed, is Hudson.
http://192.77.55.3/opengov/champlain/index.html
Champlain is a WWW-based keyword search and retrieval tool that covers
Canadian federal, provincial and municipal government sites and Canadian legal
sites on the Internet. It uses the Harvest Broker System to search WWW,
Gopher, and FTP resources (over 170 government Internet servers with over
73,000 "objects" and 23 legal Internet servers with over 2,000 "objects"
according to the latest statistics posted at the site). 9
Several keyword searching options can be selected, such as case
sensitivity or
inexact spelling, etc. Help for formulating queries is available, with clear
examples. The results can be displayed with descriptions, matched lines from
the text, in a "verbose" or summarized form. The description is just one
line, but is usually very helpful in determining the source or type of
document retrieved. The URL and path are always displayed as well, and as
hypertext links, provide direct access to the documents.
At the moment, Champlain's display of retrieved sets (with no obvious
duplication, and with the extra details given that make it possible to judge
the usefulness of a document without having to retrieve the full text) beats
GIFT's display. However, both systems will be changing. Champlain has been
working well since it came out in April 1995, although it it can be very slow.
According to Tyson Macaulay, who developed and manages Champlain for Network
Services Development Group, Industry Canada, they are looking at making some
changes soon, that will improve Champlain's search capability and increase its
retrieval speed. Another improvement I could suggest would be updating more
frequently than on a monthly basis. A month on the Internet is a long time.
It is very helpful though, that the last update is clearly stated.
(No public URL available yet)
Hudson is intended to be a free public service offering enhanced,
single-source, "blue
pages" directory-style information for all levels of government
in Canada. It should help to locate government services, programs, and
employees. Hudson will likely become the main government telephone directory
for all Canadian governments. It is being built around the new WHOIS++
protocol which allows for a distributed system of individual directory
services from all over the Internet to be "connected" for the user's search,
but remain under the control of the originators, who will be responsible for
updating and maintaining the information.
There is no public address yet for Hudson. Demonstration sites have
been set
up within Industry Canada for testing. As it looks now, after entering a
name, the display includes a picture of the person, fax and phone numbers,
Internet address, position, organization, branch, mailing address, URL, and
details on the server. Hudson's pilot launch is scheduled for spring 1996,
and the official launch in fall 1996. As with the GIFT project, government
participation in Hudson is purely voluntary.
A great deal of work has been done over the last two years, and is still
underway, both visibly, on the Internet, and not so visibly, behind the
scenes, to make locating government information on the Internet easier and
more efficient.
There is still much that needs to be done. For example, something needs
to be
done to ensure that older electronic publications can be located as well as
current ones. As most of these systems are being set up now, the originating
institution is responsible for maintaining and updating their information.
The GIFT project suggests a solution by extending that responsibility to
storing the information as well. In the text titled "GENet Contains a Real
GIFT" by GIFT Project Manager, Oliver Javanpour, he states that each version
of a document is to be considered another new document and identified
uniquely, and "after one year, most documents are placed on a tape and
warehoused and the ISBN numbers should facilitate locating these documents
years from now." 12 I wonder how accessible the
documents will be in this kind of
"warehousing", or if libraries should be more involved. The National Library
of Canada is acquiring electronic documents, but some people argue the NLC
cannot collect them all. Even if it does, some form of depository system for
electronic publications may be required. A certain amount of duplication may
be appropriate. Certainly, finding government information for historical
research will be impossible if government institutions are allowed to update,
revise, or replace existing electronic documents directly without some form of
requirement to archive, store or duplicate each version for storage elsewhere.
Another issue is the fact that all of the new projects are being
instituted on
a voluntary basis. Although it seems that government institutions have no
reason not to participate, it remains to be seen if this method will result in
comprehensive databases and directories. (It has not in the print world.)
Obviously, this has consequences affecting the ease with which information
can be located.
The GIFT initiative, in which all electronic documents are to be
described and
identified in a standard way, is perhaps the most important of all the
projects. Conforming to international standards, it also represents an
enhancement of the internationally recognized GILS (Government Information
Locator System) standard, and could lead the way to better management of
electronic government information, not just in Canada, but throughout the
world. Special attention, however, has to be given to the metadata
creation (in particular the subject descriptors), or any improvements
gained in "information management" may not be reflected in easier or
improved access.
Anita Cannon, "Finding Canadian Government Information on the Internet:
A Look at Four Principal Sites and Their Initiatives," Government
Information in Canada/Information gouvernementale au Canada, Vol.
2, no. 4.1 (spring 1996).
<URL: http://www.usask.ca/library/gic/v2n4/cannon2/cannon2.html>
[3] Canada Communications Group - Publishing.
Weekly Checklist of Canadian Government Publications.
Also available:
http://www.gc.ca/multigate.html
[4] See:
http://canada.gc.ca/whats/prgtis.html
[5] See:
http://www.gc.ca/multigate.html
[6] See:
http://www.gc.ca/comments/crtc_e.html#CONTENT
[7] Document found at the GIFT site by a keyword search
-- no URL available.
[8]
http://www.gc.ca/comments/crtc_e.html#CRTC
[9] See:
http://www.intergov.gc.ca/ABOUTIG/aboutig.html#CONTENT
[10] See:
[11] See:
http://192.77.55.3/opengov/champlain/stats.html
and
http://192.77.55.3/opengov/champlain/law-stats.html
[12] (Retrieved from the GIFT site by a keyword
search -- no URL available.
This article reviews four Internet gateway sites provided by the Government
of Canada and some of the finding tools being developed at these sites: a)
The Government of Canada Primary Internet Site, b) Intergovernmental On-Line
Information Kiosk (an intergovernmental project), c) National Library of
Canada's WWW Site, and d) Open Government Project. The search and retrieval
tools mentioned include GIFT (Government Information Finder Technology),
Champlain, and Hudson.
Introduction
The Government of Canada Primary Internet Site
GIFT: Government Information Finder Technology
Intergovernmental On-Line Information Kiosk
National Library of Canada
Open Government Project
Champlain: Canadian Information Explorer
Hudson
Conclusion
Notes
[1] May be cited as/On peut citer comme suit:
Anita Cannon
Reference Librarian
Mount Allison University
Sackville, New Brunswick
E0A 3C0
acannon@mta.ca