Developing imaged KOS with the FRSAD model: a
conceptual methodology
Simone Bastos Vieira, sbastosvieira@gmail.com
Faculty of Information Sciences, University of Brasilia, Brazil
Marcilio De Brito, mdebrito@unb.br
Faculty of Information Sciences, University of Brasilia, Brazil
Widad Mustafa El Hadi, widad.mustafa@univ-lille3.fr
GERIICO Laboratory, University of Lille 3, France
Maja Žumer, Maja.Zumer@ff.uni-lj.si
Faculty of Arts, University of Ljubljana, Ljubljana, Slovenia.
Abstract
This proposal presents the methodology of indexing with images suggested by De Brito
and Caribé (2015). The imagetic model is used as a compatible mechanism with
FRSAD for a global information share and use of subject data, both within the library
sector and beyond. The conceptual model of imagetic indexing shows how images are
related to topics and ‘key-images’ are interpreted as nomens to implement the FRSAD
model. Indexing with images consists of using images instead of key-words or
descriptors, to represent and organize information. Implementing the imaged
navigation in OPACs denotes multiple advantages derived from this rethinking the
OPAC anew, since we are looking forward to sharing concepts within the subject
authority data. Images, carrying linguistic objects, permeate inter-social and cultural
concepts. In practice it includes translated metadata, symmetrical multilingual
thesaurus, or any traditional indexing tools. iOPAC embodies efforts focused on
conceptual levels as expected from librarians. Imaged interfaces are more intuitive
since users do not need specific training for information retrieval, offering easier
comprehension of indexing codes, larger conceptual portability of descriptors (as
images), and a better interoperability between discourse codes and indexing
competences affecting positively social and cultural interoperability. The imagetic
methodology deploys R&D fields for more suitable interfaces taking into consideration
users with specific needs such as deafness and illiteracy. This methodology arouse
questions about the paradigms of the primacy of orality in information systems and
pave the way to a legitimacy of multiple perspectives in document indexing by
suggesting a more universal communication system based on images.
Interdisciplinarity in neurosciences, linguistics and information sciences would be
desirable competencies for further investigations about he nature of cognitive processes
in information organization and classification while developing assistive KOS for
individuals with communication problems, such autism and deafness.
Keywords: Knowledge representation; Semiotics; Indexing with images; Imaged
OPACs; FRSAD; Social interoperability
2
1. Introduction
Digital libraries (DL) concept is evolving to an innovative understanding of
institutions from a virtual realm that embodies in priority library concepts rather then
technological apparatus, DL are intentionally closer to its users despite being at
distance whilst adhering to principles of quality in services. That means DL is a set of
characteristics identifying documentary organizations attempting to “associate
immaterial with material (for example the OPAC1), local with distant (open/close/semi-
open client-server architecture) and visible with invisible (semantic data on the
bibliographical notes)” (Papy, 2016, 13).
The power of digitization causes scientific, educational, economic and cultural
communities to change their modes of accessing, sharing and disseminating
knowledge. Besides, complexity and uncertainty from compulsive modernisation drive
people to rely on information to perform their professional activities or to simply
exercise their citizenship. In this social flow, the world became a great consumer of
images because they are more effective in mass communication than written-only
documents (Social Bakers, 2014; Asthon, 2015).
Semiotics shows that image is a communication language and people have natural
competencies to assimilate linguistic properties from images to engage in a
communication process, which corroborates with prior linguistic theories. In fact, the
grammar codes that govern imaged communication are easily understood by a larger
public, although structurally more complex and less explicit.
Indexing with images consists of using images instead of key-words or descriptors,
to represent and organize information. Profundity in discourse is a characteristics of the
written language, yet its representation is only possible with elements of cognition. It is
a complex process to make use of words to represent knowledge. Yet, text contents
have been embodied by textual elements themselves. Admitting the multidimensional
characteristics of images to broadcast messages, this work proposes a different point of
view in the communication perspective. It certainly means a shift in knowledge
organization systems (KOS) and leads towards the achievement of social and cultural
interoperability as defined by Mustafa El Hadi (2015).
For example, in terms of knowledge organisation, the brain of a born deaf individual
(Marschark et al., 2000; Mcevoy et al., 2004) does not create the same connexions as
hearing individuals. They generate alternative logical networks to understand and
interact with the textual (oral and written) society. However, concerning the pervasive
imaged world, deaf people attest far more awareness of images than hearing people.
These observations lead to significant understandings of image perceptions. Evidences
are that accuracy in identification of image details is related to the ability of decoding
imaged messages. Because messages are codes, frequently converted to textual
systems, the decoding methods for imaged messages are the key for interoperability
between written and imaged communication. Regardless of the predominant
competence of deaf people in recognising images, our society still cannot give them the
freedom to evolve independently in our written world. There is potentially a bridge
1
Online Public Access Catalog
2
Work presented at ISKO-BRAZIL CONFERENCE (2015)
3
Latin, boundary marker, limit — more at term. First Known Use: circa 1617. Source: Merriam-Webster's
Learner's Dictionary
4
Nieuwenhuysen, P. (2013). Search by Image through the WWW: an additional tool for information
3
connecting messages from these distinct universes, establishing an authentic two-way
communication process.
What hinders semantic and consequently cultural interoperability is that the degree
of success that can be achieved in the integration of multiple knowledge representation
systems or knowledge organization schemes is constrained by limitations of the
universality of human conceptual systems. In contrast, considering that images are
powerful communication tools, imaged KOS are potentially valuable "intercultural"
interfaces for semantic interoperability which is one of the core elements towards
cultural interoperability. In this sense, we support the idea that imaged KOS support
translatability and additionally promote navigation within social and cultural diversity
through their iconic interfaces. We will develop in the following sections the concept
of the imaged online public catalogue (iOPAC)2 as a KOS for accessibility in libraries,
archives and museums.
To go further in developing the imaged KOS we will consider implementing it as a
user-focused mechanism compatible with FRSAD (Functional Requirements for
Subject Authority Data). FRSAD supports the idea that a work has subject(s) (thema),
and a thema has one or more appellations nomen. While associated with topics, images
can be interpreted as nomens within the FRSAD conceptual model.
2. Functional Requirements for Subject Authority Data (FRSAD)
IFLA proposed in 2010 a new bibliographic infrastructure to support global sharing
and reuse of subject authority data, the FRSAD model. Aboutness is the FRSAD focus
to provide a clearly defined, structured frame of reference for relating the data that are
recorded in subject records. Žumer, Zeng and Salaba (2012), show how this structure is
tailored to meet the needs of the users of these records and to assist in an assessment of
the potential need for a global information share and use of subject data both within the
library sector and beyond.
In FRSAD model a work has subject(s) thema, and a thema has appellation(s)
nomen. Nomen is any sign or combination of signs (images inclusive). FRSAD presents
four subject authority data user tasks. To find an entity (thema or nomen) or set of
entities corresponding to stated criteria. To identify an entity (thema or nomen) based
on certain attributes or characteristics. To select an entity (thema or nomen). And to
explore any relationships between entities (thema or nomen), correlations to other
subject vocabularies and structure of a subject domain. A nomen can be human-
readable or machine-readable.
2
Work presented at ISKO-BRAZIL CONFERENCE (2015)
4
Figure 1- FRSAD Relationships
Source: adapted from Riva and Žumer (2015)
FRSAD Entity-Relationship conceptual model (Riva and Žumer, 2015; Zeng, Žumer
and Salaba 2011,15) postulates that “has appellation/ is appellation of” relationship is
in general many-to-many. A thema has one or more nomens and there may be a nomen
referring to more than one thema. In addition the diagram above shows how nomens
are related to images (key-image descriptors) and traditional relationships with
thesaurus. Images can be tagged with entries from thesaurus (Gheorghita, 2011,
Aitchison and Clarke, 2004) or with external social tags, such as from folksonomies.
As a result the relational model illustrates that key-images (tailored made) are tagged
as well as original images explaining inheritance from concepts embedded in images,
what ensures wider semantic fields for imaged descriptors and sense integrity
throughout the indexing process. Moreover, the same integrity improves KOS
functions and information retrieval quality in interoperability perspectives.
By following interoperable web design principles, such as the use of RDF and XML,
this proposal sought to create an imagetic model that could be integrated into an OPAC
system potentially supporting a variety of uses, including query expansion in search,
resolving ambiguous subject use in resources, and promoting serendipitous discovery
of new links among and between documents. In short, to access, convert, or reuse web
content from a variety of domain-specific and general resources.
3. The underlying semiotic contribution
In practice, indexing as it is known is a translation of lexical units drawn from
language, or a syntactic translation reflecting the relationships between parts of the
speech, the ones describing content, the descriptors. It aims to represent the objects
(concepts) that the document is talking about, or in other words, what is said in the
speaker’s message. Although words are used to index content, the words from a natural
language or from dictionary (morphemes) point only to their signifiers, not to referents.
They do not have these characteristics of designing objects.
5
Words in terminology3 introduce a notion of a boundary or “terminus” in Latin that
gave the word “term”. The French linguist Michel Le Guern (1989) explains that in the
lexicon as in terminology there are words on both sides, but they are not the same
words. The object “word” from the lexicon is a distinct reality, the lexicon processes
words disconnected from the objects, while in terminology the words are connected to
things. Words in lexicography are considered to be nouns but in reality they are
predicates. They talk about qualities, not about substances; they refer to proprieties, not
to substances; to qualities, not to objects. The presence of an object calls for a term to
be admitted into the discourse.
For a long time the advantages of using images to transmit and receive messages
have been investigated in semiotics. Advances on information technologies encourage
important considerations about the information media. Literacy is essential, nonetheless
we all read differently. The registered information is not only obtained from text, image
has (re)conquered its place in communication. Interpreting imaged signs is as to reopen
the gates of ancient temples reading stories trough cultures and ages.
In his pragmatism and logic, Charles Sanders Peirce (1839-1914) assumed that there
is no immediate (without signs) perception of reality, so in the perception process
everything is a sign including thoughts (Nöth, 2012). He claims that a sign consists of
three elements, one of them is the sign, the second is an object to which a sign refers
and the third, the most significant, is an interpretant (Lefebvre, 2007). In its simplified
triad form of icon, index and symbol, Peirce’s logic explains that an icon represents an
acquired experience from the past, a reminder from that experience, a portrait of a
concluded moment. An index apprehends its experience from the present, it points to a
thing, without giving any information about it. The symbol is the real fact that will be
experimented in the future, at the very moment when the message arrives to the
receiver and the formulated intensions are to be re-established.
For Peirce, language and image compose the signs whose functions are embedded in
the relations between the icon, the index and the symbol. When the roles of those signs
are in balance, there is fulfilment. Peirce’s theory is largely employed for image
description purposes and in archival practices. The focus is on the description of an
‘image-document’ for further retrieval.
Instead of describing images, we are interested in using them as a sign, replacing
key-words with images for content representation. This imaged representation
presupposes a different navigation approach for the KOS and a subjacent (lying under
or below Peirce’s theory) theory to support the whole. For Peirce each object is in
relation with a number of other objects from the same universe, directly or indirectly,
in such a path that each element carries within itself an undetermined indicial potential.
How then a sign can represent an object and reveal it itself? The iconic representation
is a quality that emerges from the object in order to identify it as it is. Signs leading to
inferences over true realities are the ones that make possible indexing with images
principles as explained by De Brito and Caribé (2015).
The emerging ‘key-image’ concept means substitution or equivalence with the key-
word concept commonly used to describe subject contents. The procedure of
intentional construction of images of this model is inspired from Jacques Bertin’s
(apud Dantier, 2008; Bertin, 1970) works in which he demonstrates that image
composition follows the rules of linguistic semiotic. For this author, every thought is
3
Latin, boundary marker, limit — more at term. First Known Use: circa 1617. Source: Merriam-Webster's
Learner's Dictionary
6
expressed throughout a system of signs imitating a natural codification. The verbal
language is a code of audible signs, the writings of a language are another kind of code,
and so it is a graphic representation. So, if a graphic representation is a transcription of
information from a graphic system of signs, then it has to be considered semiotic
matter.
For Bertin (1970) a graphic representation might have three basic functions: register,
communicate and process information. Indexing with images anticipates a moment of
lecture and another of creation. It results in building an image (a chart) able to
communicate a thematic message corresponding to the document’s contents. In
comparison to the key-words, the result of indexing with images is not a simple
selection of images subjacent one to another, but a chart of significative images
intentionally composed to become the key-image of the document. These indexing
images are made of a set of iconic, indicial and symbolic proprieties (transmitted or
inherited) to represent the document. This new composed image, gathering multiple
semiotic traits, has itself a new symbolic interpretation.
4. The imaged KOS
The KOS we describe here is not an OPAC but a model for indexing by images to
support an OPAC based on a semiotic approach. Designing OPACs, using the imaged
indexing initiative, will involve not only IT solutions but also giving answers to
epistemological questions inherent to the nature of images. The procedure of using
images to describe documents is not free of impacts. Besides, Papy (2016, 57) reminds
us that it is imperative to meet the users’ needs by rethinking the design of devices and
stopping privileging the technological orientation.
Structurally, while looking for a specific document using the iOPAC interface, at
user’s behaviour first tends to recognize the conceptual relationship between images
and objects (image-subject/document-subject), then he uses this cognitive mechanism
to find the object (a document) or set of objects he is looking for. In another case, at
user has an information need, but ignores that there is an object (or a set of objects) that
could respond to his demand. He tries first to match needs to images, and then to verify
inside the collection of documents if there are connexions of some kind, repeating the
first case above.
Implementing the imaged navigation in OPACs denotes multiple advantages derived
from this. The iOPAC has a greater visual attraction pushing users towards the
catalogue; a more intuitive comprehension of indexing codes, a larger conceptual
portability of descriptors (as images), and better interoperability between discourse
codes and indexing competences affecting positively social and cultural
interoperability.
Applying the imaged concept to nomen in the FRSAD model is rethinking the
catalogue anew, since we are looking forward to sharing concepts within the subject
authority data. This happens when images, carrying linguistic objects, permeate inter-
social and cultural concepts. In practice it includes translated metadata, symmetrical
multilingual thesaurus, or any traditional indexing tools.
Thinking interfaces for the iOPAC at our ongoing project, the preliminary proposal
(Figure 2) is designed for interoperability and portability devices. In the graphic
interface, the user’s query progress in a faceted hierarchic sequence of images choices.
The cascade navigation rolls down to sub-classes inside the catalogue until reaching the
7
most profound hierarchical level, the one in which the user finds the wanted
documents.
The undertaking implementation shown in Figure 2, intends to connect the
thesaurus to the indexing methodology of key-image. It is thought to be a hierarchical
navigation throughout classes of a given classification instrument (Dewey
Classification, for example), each level associated to sets of representative key-images
of the library resources. In that way the navigation is faceted because it evolves
horizontally between choices of classes and vertically among subclasses. Each level of
subsequent prospection is assisted by sets of key-images to refine the documentary
retrieval process.
The navigation interface (Figure 2) also uses images (Library of Congress
cataloguing icons, Figure 3) instead of the common written or numeric representation
to prepare the users to the imaged representation of resources and the imaged retrieval
method of matching images to contents. From the experience of teaching deaf people in
photographic communication, they need to be coached while learning or dealing with
our common social codes (Gurgel, De Brito and Silva, 2016), in that way we expect
that LCC icons will guide the users to become more confident while exploiting the
imaged retrieval system.
Figure 2- The faceted interface for the iOPAC
The facet display represents personal and corporate images, geographic maps, time
periods, topics and others. The facet defines the interest domains; the navigation
progresses by examining representative domain material or collecting images that
reflect domain interests, then sorting images into homogeneous, mutually exclusive
groupings (facets). Facets are structured in hierarchical order to coalesce synonyms,
identify misaligned signs and gaps in the system.
The Library of Congress Subject Headings has been assigned to a simplified list of
signs making the classification schema easier for users to understand, apply, and reuse.
Any valid set of Library of Congress’ icons (Figure 3) can be converted to subject
headings. The great advantage on using LCC icons, for example, is that the image set is
homogeneous and intended to be intuitive.
8
Figure 3– Library of Congress catalogue icons
https://commons.wikimedia.org/wiki/File:The_Library_of_Congress_Classification_System.png
As shown in Figure 4, a book could be located using a traditional (written) query or
an imaged one. In both cases the search would be naturally improved by the
iconographic support of the key-image, meaning references to the book cover, to
France, to tree grammar representations, and to the authors’ image, all references
absent in the document description as shown below.
Figure 4- Book record simulation with key-image indexing
Source: Simulation study on Library of Congress’ OPAC response
(https://catalog.loc.gov/index.html)
9
The iOPAC embodies efforts focused on conceptual levels as expected from
librarians. Also, model implementations have encountered challenges during its
validation regarding methodology of mapping concepts in images or establishing
conceptual relationships among subjects and classification systems. The prototype of a
beta system is to be developed to test the model.
Research projects such as Nieuwenhuysen (2013, 2015) in which a query does not
consist of text but of an image file45 or that of Ménard (2009) could also give a larger
conceptual portability of descriptors (as images). It sought to identify the differences
from the terminological, perceptual and structural points of view, in a multilingual
environment, between indexing a set of images representing objects using either a
controlled vocabulary or a free or uncontrolled vocabulary approach. Yet, these are
nothing but imminent issues for future research.
The digital document in digital libraries, reminded Papy (2016, 103-104) introduces
combinatorial diversity of the content, redefines the relationship between content and
usage. “The form – the presentation which cannot simply be reduced to the page setting
or the formatting of the characters – carries with it a diverse set of usage semiotic
values”.
As Papy (2016) observed, the flexibility of the interoperability of Web applications
borne by XML technologies and generalization of dynamic web, the possibilities of
adapting the final form of an XML document, building it or transforming it are,
essentially limitless. Personalized virtual documents (PVDs) which are largely
composed of all kinds of fragments, extracted from different information systems,
possibly hosted on distant machines, have no predefined size or form. “Digital libraries
lend themselves to the extraction of such fragments: a video resource extracted from an
AV resource portal can be integrated into a resource with articles from different
reviews portals and bibliographical elements from online catalogs” (Papy, 2016, 93).
5. Final considerations
Indexing with images brings us to reconsider the current paradigms about using key-
words to describe document contents. The key-images announce a legitimate approach
to index documents with multiple perspectives in technical, professional and social
areas. We can perceive changes in documentary retrieval fields, when enhanced with
universal KOS based on imaged communication, and no longer contained by a specific
written language. This contribution to knowledge representation is supported by
semiotic theories and presents a new approach for documentation which needs to be
tested in large scale of documents and users.
Web search engines may use these alternatives of indexing techniques to support
KOS performance. A broader area for interface development is now available with
effective benefices for handicapped users, such as deaf people, groups with functional
illiteracy in general or in the sense of multicultural interoperability. In short, the use of
4
Nieuwenhuysen, P. (2013). Search by Image through the WWW: an additional tool for information
retrieval. In full text published in proceedings of the international conference on Asia-Pacific Library and
Information Education and Practices= A-LIEP (p. 38).
5
Nieuwenhuysen, Paul (2015) Search by Image Through the Internet: An Additional Method to Find
Information. In: International Conference on Libraries (ICOL) 2015, 25-26 August 2015, Vistana Hotel,
Pulau Pinang.
10
images to create bonds between people and documents meet KOS challenges in a
prosperous scientific ground.
The Imaged model, in addition, can be implemented as a user-focused mechanism
compatible with FRSAD. We believe that the Imaged methodology can offer new
possibilities for considering semantic and cultural interoperability when using OPACs.
The iOPAC is proposed as a product, a service, an interface and also an assistive
technology for social interoperability. The iOPAC is our ongoing project that a future
prototype beta system is to be tested.
6. References
AITCHISON, Jean and CLARKE, Stella Dextre (2004). The Thesaurus: A Historical
Viewpoint, with a Look to the Future, Cataloging & Classification Quarterly, Vol. 37,
N°. 3/4 2004, pp. 5-22.
ASTHON, Danny (2015). 10 reasons why you should care about visual content
marketing. Blog. Publicado em 19 de fevereiro, 2015. Disponível em:
. Accessed : May 2015.
BERTIN, Jacques (1970). La graphique. Communications, v. 15, n. 1, p. 169–185.
CARIBÉ, Rita de C.V and DE BRITO, Marcílio (2015). Indexação por imagens: via
OPACs imagéticos. In: III Congresso Brasileiro de Organização e Representação do
Conhecimento. Marília - SP: ISKO-Brasil. p. 425-448.
DANTIER, Bernard (2008). La représentation et l’étude visuelles des informations. In.:
BERTIN, Jacques. Semiologie graphique, les diagrammes, les reseaux, les cartes. 4.ed.
Paris. Ehess, 32 p.
DE BRITO, Marcílio and CARIBÉ, Rita de C.V (2015). Princípios da indexação por
imagens. In: XVI Encontro Nacional de Pesquisa em Ciência da Informação (XVI
ENANCIB), João Pessoa p. 1-2
GEMBERLING, Ted (2016). FRSAD, Semiotics, and FRBR-LRM. Cataloging &
Classification Quarterly 54, no. 2: 136. Publisher Provided Full Text Searching File,
GHEORGHITA, Inga (2011). Methodologie de construction automatique du thesaurus
pour l’indexation et la recherche des images. RECITAL, n. 2, p. 221–228, Jun.
Disponível em: . Accessed : May 2015.
GURGEL, I.; DE BRITO, M. ; SILVA, R. C. J. (2016). Projeto tecnológico e
sustentável de inclusão social: Rede de bibliotecas com acessibilidade para estudantes
surdos In: I Congresso Internacional de Educação Especial e Inclusiva, 2016. Marília -
SP. I CIEEI & XIII JEE 18 a 20 de Maio de 2016 - Marília - São Paulo. Marília - SP:
Fundepe.
IFLA (2000). Guidelines for Library Services to Deaf People. 2nd edition. Hague:
IFLA
11
LE GUERN, Michel (1989). Sur les relations entre terminologie e lexique. META
Journal de traducteurs. Montréal.
LEFEBVRE, Martin (2007). The art of pointing: on Peirce, indexicality, and
photographic images. Photography Theory, p. 220-244.
MARSCHARK, M; GREEN, V; HINDMARSH, G; WALKER, S. (2000).
Understanding theory of mind in children who are deaf. Journal of Child Psychology
and Psychiatry 41 (08), 1067-1073.
MCEVOY, C; CONVERTINO, C; MASTELLER, A; MARSCHARK, M (2004).
Organization and use of the mental lexicon by deaf and hearing individuals. American
Annals of the Deaf, 149 (1), 51-61
MÉNARD, Elaine (2009). Images: indexing for accessibility in a multi-lingual
environment -- challenges and perspectives. Indexer, Jun 27( 2), 70-76.
MUSTAFA EL HADI, Widad (2015). Cultural Interoperability and Knowledge
Organization Systems. Keynote address, University of St Paulo at Marilia, Brazil,
Proceedings of the 3rd Brazilian ISKO-Conference.
NIEUWENHUYSEN, P. (2013). Search by Image through the WWW: an additional
tool for information retrieval. In: The international conference on Asia-Pacific Library
and Information Education and Practices A-LIEP (p. 38).
NIEUWENHUYSEN, Paul (2015) Search by Image Through the Internet: An
Additional Method to Find Information. In: International Conference on Libraries
(ICOL) 2015, 25-26 August 2015, Pulau Penang. Malaysia.
NÖTH, Winfried (2012). Charles S. Peirce's theory of information: a theory of the
growth of symbols and of knowledge. Cybern. Hum. Knowing 19 (1e2), 137 e 161.
PAPY, Fabrice (2016). Digital libraries: interoperability and uses. Oxford: Elsevier.
146p.
RIVA, P. and ŽUMER, M (2015). Introducing the FRBR Library Reference Model. In:
IFLA WLIC 2015 - Cape Town, South Africa in Session 207 - Cataloguing. 7p.
SOCIAL BAKERS (2014). Photos cluttering your facebook feed? here’s why pictures
account for majority of brand facebook posts. April 21. Advertising & Marketing.
Available at: Accessed : May 2015.
TUFTE, Edward R (1998). Envisioning information. Connecticut, USA: Graphics
Press. 126 p.
ZENG, Marcia Lei, Žumer, Maja and Salaba, Athena. (2011). Functional requirements
for subject authority data (FRSAD): a conceptual model. Berlin: De Gruyter Saur.
ŽUMER, Maja; SALABA, Athena and ZENG, Marcia Lei (2007). Functional
Requirements for Subject Authority Records (FRSAR): A Conceptual Model of
Aboutness. In: Proceedings of the 10th International Conference on Asian Digital
Libraries (ICADL ), Hanoi, Vietnam, December 10-13, 2007. (Lecture Notes in
Computer Science Series, Volume 4822: 487-492) Berlin: Springer.
12
ŽUMER, Maja; ZENG, Marcia Lei and SALABA, Athena (2012) FRSAD: Conceptual
Modeling of Aboutness. Third Millennium Cataloguing. Santa Barbara: Libraries
Unlimited.