Knowledge Structures and the Internet: Progress and Prospects

Page created by Greg Jacobs
 
CONTINUE READING
Knowledge Structures and the Internet:
             Progress and Prospects
                                      Nancy J. Williamson

        SUMMARY. This paper analyses the development of the knowledge
        structures–Web directories, thesauri, and gateways/portals–as they are
        presented on the Internet as aids to information seeking. It identifies
        problems and suggests improvements. doi:10.1300/J104v44n03_11 [Arti-
        cle copies available for a fee from The Haworth Document Delivery Service:
        1-800-HAWORTH. E-mail address:  Web-
        site:  © 2007 by The Haworth Press, Inc. All
        rights reserved.]

        KEYWORDS. Thesauri, Web subject directories, gateways, portals,
        subject access

                                        INTRODUCTION

   Since its early and somewhat primitive beginnings, the Internet has
gone through many changes. It has grown ever larger, and search en-
gines have become increasingly sophisticated. In efforts to improve ac-

   Nancy J. Williamson is Professor Emeritus, University of Toronto, Faculty of Infor-
mation Studies, 140 St. George Street, Toronto, Canada, M5S 3G6 (E-mail: nancy.
williamson@utoronto.ca).
    [Haworth co-indexing entry note]: “Knowledge Structures and the Internet: Progress and Prospects.”
Williamson, Nancy J. Co-published simultaneously in Cataloging & Classification Quarterly (The
Haworth Information Press, an imprint of The Haworth Press, Inc.) Vol. 44, No. 3/4, 2007, pp. 329-342; and:
Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter (ed: Robert P. Holley) The Haworth Infor-
mation Press, an imprint of The Haworth Press, Inc., 2007, pp. 329-342. Single or multiple copies of this arti-
cle are available for a fee from The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. -
5:00 p.m. (EST). E-mail address: docdelivery@haworthpress.com].

                      Available online at http://ccq.haworthpress.com
                   © 2007 by The Haworth Press, Inc. All rights reserved.
                               doi:10.1300/J104v44n03_11                                                 329
330    Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

cess and to ease the search for information, old tools have been adapted
and reconfigured; and new tools are constantly being developed. The
question of how to organize Web sites so users can actually find what
they are looking for is a continuing problem. In any event, while infor-
mation systems may change to achieve greater success, they must con-
tinue to meet two fundamental requirements of information seekers: to
permit users to locate information on a subject directly and to allow
them to browse so as to familiarize themselves with a domain or to re-
fine a request.
   In the infancy of the Internet, information providers were working on
a trial and error basis. Research findings indicate that the expertise of
the providers varied in its sophistication. The medium and its potential
were often misunderstood. Information seeking methods were some-
times naïve and the access tools inadequate to the task. These findings
were confirmed in the research for two papers previously prepared–
“Knowledge Structures and the Internet” (Williamson, 1997) and “The-
sauri in the Digital Age” (Williamson, 2000). In 1997, search engines
were relatively primitive; and control over the development of the
Internet was minimal. Each information provider had his/her own ob-
jectives as to how information should be organized to accomplish easy
and productive browsing. The literature seldom addressed the intrica-
cies of structuring the data. There existed no standards or generally ac-
cepted guidelines for dealing with document content. Moreover, early
research and development focused on the societal and technological
problems rather on a concern for effective access to data. Indeed, like so
many new technological toys, the developers and many users assumed
that the entertainment value of ‘surfing the net’ was its most important
function. Research into the possibilities for organization and access had
only just begun.
   In the 9 years (2006) since that time, much has changed. The Internet
is now seen as a serious source of information in the academic, busi-
ness, and industrial communities as well as in the eyes of the ordinary
citizen. In 2006, Web designers and researchers are well into exploiting
a full range of knowledge structures and search strategies. Software
cannot solve all the problems. There is an urgent need for more user
friendly interfaces and greater emphasis on human computer interaction
to aid the user in achieving successful searches. From its early begin-
nings, much effort has been put forward in an effort to “index the
Internet.” The first approaches were rather simplistic in nature and not
always appropriate to the new medium. For example, early use of tradi-
tional classification was at a minimal level, usually shallow and some-
Part IV: Position Papers                      331

times erroneous. Currently, controlled vocabularies of various kinds
(e.g., thesauri and taxonomies) as well as other kinds of information
structures are deemed to have an important role to play. Most signifi-
cantly, endeavors to create seamless information systems in many cases
have led to the integration of the tools into the databases themselves. At
this point, major questions to be addressed are: “How well are these
tools performing their task? Is there room for improvement?” Using the
aforementioned papers as a starting place, this current research exam-
ines the access tools provided for the aid of users. Focus is on those de-
vices that can actually be viewed on the screen by users and can be
manipulated to facilitate subject searching. Emphasis is on structure and
complexity as embodied in the use of tools that foster browsing subject
domains and individual Web sites. Structure has been defined as the
bringing together and organization of information in a way that facili-
tates browsing. Included is classificatory structure as embodied in both
traditional and newly developed aids to searching.

                           METHODOLOGY

   As the first step in this investigation, a literature search was conducted
to locate relevant research and publications since 1997. There are numer-
ous devices to aid the user such as ‘more like this,” hyperlinking, site
maps, etc. Since the list is long, three major tools were selected for more
intense study–Web directories (taxonomies), thesauri, and gateways/
portals. Hyperlinking and mapping are discussed in conjunction with
these three. From the results of this search, the investigation was ex-
tended further through a critical examination of the design and structure
of relevant Web sites in each of the categories. Since much of the litera-
ture was published prior to 2003, an accurate picture could be obtained
only by viewing current sites themselves.
   Some of the sites viewed in 2000 (Williamson) were revisited to see
if changes had been made, and new sites were added. The ultimate goal
was to find answers to such questions as: Where do things stand in the
development of the browsing capabilities of the Internet? What is the
nature of the changes that have taken place? Are they changes that truly
support users in their searches? Are the new tools being used effec-
tively? Are there obvious improvements that could be made? Where
should research and development go from here? Final conclusions have
been drawn from the findings from the two areas of investigation.
332    Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

                          SEARCH ENGINES

   The bottom line on the Internet is the search engine, and the search is
constrained and/or supported by what that search engine has to offer.
What is the coverage of the data? What parts of the document are
searched? What search options are provided? These are things that de-
termine what can be retrieved. The other aspect of the system is its navi-
gation once the search engine has done its job. It is the navigation over
which the information seeker has partial control. The nature, strengths,
and weaknesses of search engines are not discussed here. The focus is
on the tools that aid the user to navigate Web sites through directed
searches. The analysis that follows here moves from the simplest to the
most complex.

                   WEB SUBJECT DIRECTORIES

    One of the most obvious tools in a directed search of Internet sites is
the Web subject directory. Web directories “are, in fact, a form of clas-
sification” (Gilchrist, 2003, p. 11) and are described by Gilchrist, along
with some other applications, as belonging under the more generic term
“taxonomies.” These directories fulfill the two basic requirements for
searching an information system. They permit searching on specific
terms and also allow browsing through the lists of resources. This in-
vestigation focused on subject directories of six sites that are reputed
to be “good sites”–Yahoo, Open Directory, LookSmart, Librarians’
Internet Index (LII), InfoMine, and RDN (Resource Discovery Net-
work) (Notess 2003). In each case, searches were carried out under the
terms “health” and “education.” The terms in the directories lead through
hyperlinking to Web sites which deal with each particular topic.
    In each directory, the user is offered a menu of top terms. Selecting
one of the terms, the user can move to second and succeeding levels un-
til some information or references are reached; or perhaps the opportu-
nity of using another search engine is offered. Directories are designed
by humans for each particular Web site. There are no rules or standards
for these directories, but attempts are made to use terminology and
structure suitable to the particular site (as opposed to using a standard
classification scheme). The number of top terms, the total number of
sites involved, and the nature of the domain influence the degree of divi-
sion, the number of hierarchical levels in a directory, and length of the
pathways to particular topics. For example, the Yahoo directory leading
Part IV: Position Papers                      333

to 3,000,000 + locations is much more detailed than RDN’s (Resource
Discovery Network) leading to 30,000 locations (Notess 2003). Top
terms tend to be at the domain level, but popular topics might also be top
terms. For example Librarians’ Internet Index includes the very broad
topics–Arts & Humanities, Science, and Society & Social Science as
top terms as well as terms that could be subsumed under these. At the
upper levels, terms sometimes included a brief contents note to aid the
user in making choices. Web directories are described as being hierar-
chical, but they are not hierarchical in the classical sense. In some cases,
a term may be repeated at different levels. For example in the Yahoo di-
rectory, ‘cancer’ and ‘breast cancer’ both appear at level 3; but ‘breast
cancer’ appears again at level 4. Also the categories are not “pure.” For
example, “diseases and conditions” and “news and media” are in the
same category in the Yahoo directory. Not every term relative to the
various Web sites will appear in a directory. In all the subject directories
searched, there was provision for searching on keywords as well as for
browsing.
   As might be expected, there is wide variation in the structure and use-
fulness of Web directories. The user controls the search by following a
path in the directory. A useful device is a display of the search path at
each step of the user’s search (Yahoo and Librarians’ Internet Index).
InfoMine differs from all the other directories. The first page of the site
contains the top terms in the directory. A click on the top term leads to a
search form designed for selecting search options and for inputting a
term that permits a choice of regular display or relevance ranking. As
aids in selecting search terms, browsing of LCSH, LCC, keywords, and
other indexes is also permitted. Similarly, searching ‘biological, agri-
cultural and medical sciences’ in RDN permits the use of LCSH,
MESH, and LCC.
   The large Web directories (Yahoo and Librarians’ Internet Index)
tend to be the most logically organized and most minutely divided. In
the Yahoo example, some paths went down six levels whereas RDN
provided division that was only three levels deep in the paths searched.
For the user, effective sorting of the material is important. There is need
for an increasing number of levels as the size of the database increases.
In the Yahoo directory, for example, a search for ‘mammography’ goes
through 5 levels: Health–Diseases and conditions–Cancer–Breast can-
cer–Mammography. If a directory is to continue to be useful over time,
revision and resorting of the directory needs to occur as the number of
and types of Web sites grow. There is little indication as to whether this
actually happens.
334     Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

                                 THESAURI

   While Web subject directories (taxonomies) provide useful tools to
aid in navigating the Internet, thesauri provide much more powerful
tools not only by allowing subjects to be arranged hierarchically but
also by permitting other kinds of term relationships and linkages. Con-
trolled vocabularies, such as subject headings, go back to the nineteenth
century; and the modern thesaurus emerged in the early 1960s. By the
late 1990s, it was generally assumed that there was an important role for
the thesaurus to play as a tool in online access to information (Milstead
2000). At that time, that role was still being defined. Among other
things, it was assumed that thesauri could complement full text access
by aiding users in focusing their searches by supplementing the linguis-
tic analysis of the text search engines and even by serving as one of the
tools used by the search engine for its analysis. Machine aided indexing
could make use of thesauri as a basis for easier term selection by index-
ers. Also by analogy, the principles of term relationships as applied in
thesauri should be applicable in the creation of hyperlinks in large data-
bases and on the Internet.
   Over time, a growing number of thesauri have become available
through the Internet. There are two basic types of display–static and dy-
namic. Static thesauri are displayed much as they are in printed form
and differ only in that they are available electronically. The contents can
be browsed, but there is little or no facility for moving about the list in a
dynamic way. For practical purposes, they are clumsy; it would often be
easier and more efficient to use a printed volume. This approach was
typical of initial attempts to move printed products into digital format
and was precipitated by the existing use of electronic sources to create
the printed versions. Dynamic thesauri, on the other hand, are presented
in such a way that they can be “searched” by inputting thesaurus terms
or by browsing a section of the alphabet. Some provide a list of catego-
ries as a starting place for beginning a search. Some allow Boolean
searching and begin to use hyperlinks that enable users to move from
one part of the thesaurus to another; to follow the BT, NT, and RT rela-
tionships; and to move from one type of display to another (e.g., from a
rotated display to an alphabetical display). Some are derived from a
printed product; others are newly created for the Internet and only exist
in electronic form. Some (e.g., the Cook’s Thesaurus) actually led to
documents. By 2000, there was considerable improvement in the way
thesauri were presented online. At the time, there was evidence of a
solid start in the development of online thesauri; but there was still
Part IV: Position Papers                      335

much research to be done on possible ways to enhance the display and
the use of thesauri online. There were plenty of examples of what to do
and what not to do. Many of the designs were predicated on what the de-
signers knew about databases and not always on what was known about
the behavior of users. Some were incomplete and displayed only a por-
tion of the thesaurus as an encouragement to buy the product. However,
there were healthy signs of innovation; in particular, many lists now be-
gan to lead to actual document citations–confirming the prediction that
direct linkage between a thesaurus and documents was on the way.
   All thesauri that presently exist were created, more or less, on the
basis of the current guidelines for thesaurus construction. These
guidelines, while sound in linguistic principles, are technologically
somewhat out of date. The ISO 2788 guidelines were published in 1986.
The ANSI/NISO Z39.19 guidelines, published somewhat later in 1993,
employed the same guiding principles as ISO 2788; but included a sec-
tion on screen display that recognized that “sophisticated thesaurus dis-
play and terminology” (NISO 1993, p. 25) would be appropriate for
expert searchers and indexers. Furthermore, it acknowledged the fact
that screen viewing is different from print viewing and presents difficul-
ties for the user. In addition, it alludes to the need for standards for hu-
man-computer interaction but does not refer directly to the possibility of
a dynamic thesaurus. By 1999, at a workshop sponsored by NISO on
electronic thesauri, it was recommended that a variety of flexible dis-
plays should be provided and that a new standard be established for dig-
ital thesauri. Furthermore, it was soon recognized that new guidelines
were needed. The work on new standards has now been completed.
ANSI/NISO Z39.19-2005 was published in November 2005 and is
available on the Web at http://www.niso.org/standards/resources/
Z39-19-2005.pdf (accessed 4/3/2006). BS 8723, “Structured Vocabu-
laries for Information–Guide,” parts 1 and 2, have been published in the
UK by the British Standards Institution (BSI). This guide supercedes
BS 5723:1987, the Guide to Establishment and Development of Mono-
lingual Thesauri, and is described by Leonard Will as being substan-
tially different from BS 5723. “The text has been rewritten in today’s
idiom and some additional aspects are now covered” (Will, 2006) in-
cluding facet analysis, presentation via electronic (as well as printed)
media, thesaurus functions in electronic systems, and requirements for
thesaurus management software” (Will, 2006). Three other parts yet
to come will cover “vocabularies other than thesauri, interoperation
between multiple vocabularies (with multilingual as a special case),
interoperation between vocabularies and other components of informa-
336     Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

tion storage and retrieval systems.” The hope is that “BS 8723 will pave
the way towards a corresponding revision of the international standard
ISO 2788.” Without having seen the full text of the BSI version, it can
be said that both standards support the principles laid out in previous
standards and that they are cognizant of the need to provide for elec-
tronic manipulation and display of thesauri and for their integration
with databases. Problems of interoperability will also be addressed. In
NISO Z39.19-2005 (p. 103-104), Section 11.47 sets out requirements
for “browsing within hierarchical and alphabetical displays” and the
“viewing of a term in the context of its relationships and its complete
term record from any display (through hyperlinking).”
   These guidelines should lead us into the future. But what is the state
of the thesaurus on the Internet today? Much has changed in the way
thesauri are handled. Many of the predictions made by Milstead (2000)
have come to pass. Clearly, the thesaurus has now assumed its role as a
search tool. However, there are some unfortunate links to the past. A re-
visit to some of online thesauri from the previous research found that
some of them are still available on the Internet and have never having
updated or improved. However, things have moved on; and the most
important developments are the efforts to link thesauri up with the data-
bases and the convergence of a number of databases under on one ac-
cess point. Online thesauri can be found in one or more of the following
forms (Shiri and Revie 2000): (a) simple static text format (e.g., ASFA;
NASA Thesaurus); (b) static HTML format; (c) dynamic HTML format
with fully navigable hyperlinks (e.g., MeSH; UNESCO Thesaurus;
(d) advanced visual and graphic interfaces (Plumb Design Visual The-
saurus); and (e) XML format (Virtual Hyperglossary).
   Because of the lack of standards, there are considerable differences
in: (a) the way the thesauri are accessed and (b) the provision for their
electronic manipulation. Some online versions are incomplete and are
online mostly to encourage subscriptions to the full product. Indeed, in
many cases, the only sure method of receiving updates is to subscribe.
Rarely, except in static format, can a user access the whole thesaurus at
once. With the dynamic format, the “black box” effect exists; and ac-
cess in response to input by the user of a term, partial term, or known
descriptor is either a list of descriptors or a group of related thesaurus
entries (with BT, NT, RT relationships). Most are forgiving of spelling
mistakes and respond to truncation. However, the browsing feature of
the printed thesaurus has gone the way of the card catalogue. Perhaps
this is not a problem for the searcher, but it may be for the indexer. Users
often do not know what the working version looks like without a sub-
Part IV: Position Papers                      337

scription to the tool. An essential of a good online thesaurus is hyper-
linking to permit navigation across the thesaurus, both within a
particular display and across displays. The following examples demon-
strate some of these features.
    The NASA Thesaurus is of the static type; and, yes, you can view the
whole of the alphabetical display; but it cannot be manipulated online
and is not connected to the database. The CATIE HIV/AIDS Treatment
Thesaurus is a dynamic thesaurus and is accessed through letters and
letter combinations (e.g., A–Ana, Anb-Az, B, etc.). Selection of a seg-
ment brings up an alphabetical list of terms from which a choice can be
made. The chosen term leads to a thesaurus entry with all the necessary
relationships, but that list is static and cannot be manipulated further.
Having chosen the term, the user must then go to the Web site search
page or to the library catalog to get the final information because the
thesaurus and the database are not connected.
    The UNESCO Thesaurus is a multilingual thesaurus containing
7,000 English terms, 8,600 terms in French, and 8,600 terms in Spanish.
It is easy to use and has most of the requirements for a good online the-
saurus. A search on “UNESCO” brings up the official Web site for
UNESCO documents and publications that provides a link to the the-
saurus. The thesaurus can be searched both alphabetically and hierar-
chically. An alphabetical search is initiated by inputting “a few letters”
(e.g., “cultur” for culture). The click leads the user to the relevant terms
from the permuted list. Each term includes a full thesaurus entry with
the number of documents in square brackets. BTs, NTs, and RTs are
hyperlinked to their own records so that a user can expand/narrow the
search if desired. A click on the number of documents retrieves the
records for the documents from unesdoc/bib for that topic. From that
point, the full catalogue record can be accessed if there is one; full text
can be accessed online if available; and “no full text” is indicated where
appropriate. If the hierarchical approach is chosen, the process is
slightly different. The user is asked to choose (a) a domain from a list
(e.g., Social and human sciences) and (b) a specific subject area (e.g.,
Social problems). Input again takes the user to full thesaurus entries
supported by hyperlinks. There may be more than one record. For ex-
ample, the request for “social problems” brought up thesaurus records
for “crime,” “disadvantaged groups,” and “social problems.” Clicking
the number of documents leads to records for the documents. While the
user must understand the system, it is simple and effective to use; and
the instructions are clear. It gives details of the thesaurus including
“how to browse the thesaurus.” This format stands alone and is intended
338     Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

for use as a searching thesaurus. Alphabetical access is through an al-
phabetical list of terms that leads to the thesaurus entries. While this the-
saurus would serve the user well, the indexer might have a problem as
one cannot browse the whole thesaurus. However, the working format
may have some differences. In addition, while the user can move around
in the individual format, he/she cannot move directly from the hierar-
chical format to the alphabetical format without returning to the initial
page.
   Another format now being introduced is the advanced visual and
graphic interface. This approach has great potential, but so far the exam-
ples are very simple and applied to a small group of terms. Think Map is
one example of this and, at best, might be thought to be “cute.” On the
Internet, a graphic design is presented that changes when individual
terms are clicked on. Another result is a change in the configuration
with the adding of new terms and the subtraction of others. Very little
information is given, and the user is allowed to play with it for only a
few minutes. Access to it is problematic without further information.
There may be an alphabetical list linked to the graphics, but only a sub-
scription or questioning the owner would provide answers. One would
not want to subscribe without more information. The significance of
this format is that this type of graphic design is not new. It was demon-
strated by Lauren Doyle (1961) and by Eric Johnson and Pauline
Cochrane (1998) in an experiment with the INSPEC Thesaurus. This
kind of design deserves some further thought.

                      GATEWAYS AND PORTALS

   When an extensive information system becomes extremely large, it is
difficult to locate the best materials on a domain or to find all the impor-
tant databases pertaining to that domain. At that point, there is a need to
provide one’s clients with improved access through “narrowing the fo-
cus to a super discovery tool” (ARL 2005, p. 3). Among the newest
tools coming to the aid of information seekers are gateways and portals
that provide access to Web resources and/or various databases through
one facility. They tend to be either client or subject oriented. Some
cover a vast territory, but they may also be developed by individual li-
braries and special information centers. The particular interest here is
subject gateways. More precisely defined:
Part IV: Position Papers                      339

     Subject gateways are Internet services which support systematic
     resource discovery, provide links to resources (documents, ob-
     jects, sites or services), predominantly accessible via the Internet.
     The service is based on resource description. Browsing access to
     the resources via a subject structure is an important feature. (Koch
     2000, p. 24-25)

    These are developed with care to ensure high quality. Quality-con-
trolled subject gateways are created by editors and subject specialists to
ensure a high level of quality. Completeness and balance are sought in
collection development, and a policy is developed to ensure the contents
are up to date. Quality metadata is used that should comply with an ac-
ceptable standard. Formalized content description is also recommended.
Of particular interest here is the kind of subject access. Koch (2000, p. 25)
indicates that there is a need for deeper levels of classification and that
subject/browsing structure is important. He also calls for keyword or
better controlled vocabularies (e.g., subject headings, thesauri, etc.) for
subject indexing as well as advanced search and browse access.
    As with thesauri, taxonomies, and ontologies, there is some confu-
sion over the terminology. The terms “gateways” and “portals” appear
to be used synonymously (Lancaster, 2003), and some have referred to
them as virtual libraries. One example of this is that Lancaster refers to
InfoMine and Librarians’ Internet Index as portals while they are de-
scribed elsewhere in the context of Web directories. There are a number
of important examples of these structures, and the number is growing
rapidly. Founded in 1988, the Ovid gateway provides access to elec-
tronic, scientific, and academic research information and provides for
multiple ways to carry out searches. The NLM Gateway is a Web-based
system that permits users to search simultaneously in multiple retrieval
systems at the National Library of Medicine. Users can initiate a search
from one Web interface and carry out one-stop searching in all the
NLM’s databases. One huge gateway is CSA Illumina. CSA describes
itself as a world wide information company. It sounds commercial, but
it leads to bibliographies and journals in four primary areas–natural sci-
ences, social sciences, arts and humanities, and technology. It provides
a single point of access to very large number of electronic resources.
One might argue that it is not necessarily a “quality” subject gateway,
but it all depends on the resources it leads to. It includes in its long list
ERIC and Scholar’s Portal. ERIC is having to make some concessions
to be included there. The ERIC Thesaurus has only recently become
340     Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

accessible again after being reconstructed for inclusion in the CSA
Illumina system.
   Scholars’ Portal is a system developed in 2002 by the ARL Libraries;
users of some of those libraries will be familiar with it. There may be
different configurations of the portal in different libraries. One example
of its application is the one at the University of Toronto Libraries. From
the official Web page, four approaches are possible:

  1. “Search Illumina” leads to CSA Illumina and a broad list of data-
     bases from which user can select (e.g., AGRICOLA, ERIC, LISA)
     and then be taken to a description of the resource;
  2. “Electronic journals” leads to a list of journals that can be browsed
     and accessed by letter. The journal name is hyperlinked to the ac-
     tual journal and ultimately to tables of contents that allow particu-
     lar journals to be searched;
  3. “RefWorks” permits the building of a personal bibliography of
     sources;
  4. “RACER” leads to an inter1ibrary loan function.

   However, at the University of Toronto, a user’s first encounter with
Scholars’ Portal may result from a search of such journal indexes as
LISA and ISTA where a subject search retrieves a list of references. A
particular reference leads to a title and full abstract and a direction to see
UTL (University of Toronto Libraries). Response to this command will
tell the searcher if and where full text online is available–Scholars’ Por-
tal, Haworth Press, HW Wilson, Proquest, etc. A click will bring up the
full article. If no full text is available online, the searcher is referred to
the library catalogue where it can be determined whether the book or
printed journal is available. The bibliographic and interlibrary loan
functions are also active here. The system is not finished, and there are
improvements to be made. The decision to use Scholars’ Portal is Uni-
versity of Toronto’s answer to Google.

                              CONCLUSION

   Has control of subject access to the Internet improved since 1997?
Definitely yes! There are more and better provisions for access to mate-
rials, and serious efforts are being made to single out the best materials.
This is happening at four levels–at the gateway/portal level, at the indi-
vidual database level, at the Web site level, and at the resource level.
Part IV: Position Papers                             341

Aside from the structures discussed above, lengthy pieces of text are ap-
pearing with tables of contents and outlines hypertexted to the appropri-
ate location in the text. There are indexes that look very much like book
indexes that bring together related parts internal to a text and other devices.
Most importantly, some of this might not have happened without the devel-
opment of the Web-based OPAC and the emergence of metadata. These
two developments have done much to aid the changes that have taken
place. However, there is much yet to be done. There could be further im-
provements in online thesauri to make them more useable for searching
and browsing. The gateways and portals will improve as time goes on.
Will the Internet ever be a perfect world? No. Things will become more
complex; but, in a world where everybody is an information provider,
many of them will go their own way. However, it is to be hoped that
“quality” information will rise to the top.

                                  REFERENCES
Aitchison, Jean and Clarke, Stella Dexter. (2004). “The thesaurus: a historical view-
   point with a look to the future.” Cataloging & Classification Quarterly 37 (3): 5-21.
ARL Scholars Portal Working Group. (2002). “Final report” http://www.arl.org/access/
   scholarsportal/final.html (accessed 12/12/2005).
CSA guide to discovery. http://illumina.scholarsportal.info/ (accessed 2/6/2006).
Doyle, Lauren B. (1961) “Semantic road maps for literature searchers,” Journal of the
   Association for Computing Machinery 8 (4): 553-578.
Gilchrist, Alan. (2003). “Thesauri, taxonomies and ontologies–an etymological note.”
   Journal of Documentation 39 (1): 7-17.
Garshol, Lars Marius. (2004). “Metadata? Thesauri? Taxonomies? Topic maps! Mak-
   ing sense of it all.” Journal of Information Sciences 30 (4): 378-391.
Gullikson, Shelley, Blades, Ruth, Bragdon, Marc, McKibbon, Shelley, Sparling, Marnie
   and Elaine G. Toms. (1999). “The impact of information architecture on the aca-
   demic Web site usability.” The Electronic Library 10 (5): 293-304.
International Organization for Standardization. (1986) Guidelines for the Establish-
   ment and Development of Monolingual Thesauri. ISO2788. 2d .ed. Geneva: ISO.
Johnson, Eric H. And Cochrane, Pauline A. (1998). “Hypertextual interface for a
   searcher’s thesaurus.” http://www.csdl.tamu.edu/DL95/papers/johncoch/johncoch.
   html (accessed 4/3/06).
Koch, Traugott. (2000). “Quality-controlled subject gateways: definitions, typologies,
   empirical review.” Online Information Review 24 (1): 24-34.
Lancaster, F.W. (2003). “Portals.” In Lancaster, F.W. Indexing and Abstracting: The-
   ory and Practice. 3rd ed. Champaign, Il: University of Illinois, School of Library
   and Information Science. pp. 352-354.
Milstead, Jessica. (2000). “Invisible thesauri: the year 2000.” Online &CDROM Re-
   view 19 (2): 93-94.
342      Cataloger, Editor, and Scholar: Essays in Honor of Ruth C. Carter

Notess, Greg R. (2003). “Internet subject directories.” http://www.searchengineshowdown.
   com/dir/ (accessed 4/3/06).
Renardus. “Gateways defined.” http://www.renardus.org/about_us/subject_gateways.
   html (accessed 10/20/2005).
Shiri, Ali Asghar and Crawford Revie. (2000). “Thesauri on the Web: current develop-
   ments and trends.” Online Information Review 24 (4): 273-279.
Will, Leonard. [e-mail received 31/01/06].
Williamson, Nancy J. (1997). “Knowledge structures and the internet.” In Knowledge
   Organization for Information Retrieval: Proceedings of the Sixth International
   Study Conference on Classification Research. Held at University College London,
   16-18 June 1997. The Hague, Netherlands: International Federation for Information
   and Documentation. pp. 23-27.
Williamson, Nancy J. (2000). “Thesauri in the digital age: stability and dynamism in
   their development and use. In Dynamism and Stability in Knowledge Organization:
   Proceedings of the Sixth International ISKO Conference 10-13 July 2000,Toronto,
   Canada. W«rzburg: ERGON Verlag. pp. 268-274.
Zeng, Marcia Lei and Chan, Lois Mai. (2004). “Trends and issues in establishing
   interoperability among knowledge organization systems.” Journal of the American
   Society for Information Science and Technology, 55 (5): 377-395.

                          doi:10.1300/J104v44n03_11
You can also read