=Paper=
{{Paper
|id=Vol-538/paper-8
|storemode=property
|title=Bringing the "Thesaurus for Economics" on to the Web of Linked Data
|pdfUrl=https://ceur-ws.org/Vol-538/ldow2009_paper7.pdf
|volume=Vol-538
|dblpUrl=https://dblp.org/rec/conf/www/Neubert09
}}
==Bringing the "Thesaurus for Economics" on to the Web of Linked Data==
https://ceur-ws.org/Vol-538/ldow2009_paper7.pdf
Bringing the “Thesaurus for Economics”
on to the Web of Linked Data
Joachim Neubert
German National Library of Economics (ZBW)
Leibniz Information Centre for Economics
Neuer Jungfernstieg 21
20347 Hamburg, Germany
j.neubert@zbw.eu
ABSTRACT The original goal of the current project was to make the STW
Thesauri are possible building blocks of a web of linked data. As available on the ZBW web site. The website should offer topical
DBpedia for large data sets in general, specialized thesauri could entry points to the library catalog ECONIS 1 and other ZBW data-
be useful as interlinking hubs for professional communities – if bases with a straightforward navigation within the thesaurus and
they are available on the linked data web. The paper describes the towards the retrieval of books and articles, according to the “fol-
conversion of a large economics thesaurus to RDF/SKOS, using low-your-nose” [13] principle. An automatic generation mecha-
the enhancement mechanisms of SKOS to dispose some non- nism which makes it easy to publish new versions of the thesaurus
standard features of this thesaurus. The deployment, using RDFa was also a requirement.
pages, and the interlinking with other resources, namely a library We did not want to integrate the generation of the web pages into
catalog, and an experimental mapping to DBpedia are presented. the thesaurus maintenance application, but aimed for decoupling
For information retrieval support, a SPARQL query facility uses it through an intermediate text file format. We wanted this format
the data for building a thesaurus-backed terminology web service. also to be appropriate for publishing the thesaurus as a whole, in
order to support subsequent use by others (in ways we would not
Categories and Subject Descriptors be able to predict). Encouraging secondary use within the eco-
H.3.1 [Content Analysis and Indexing] Thesauruses; I.2.4 nomics community had been a primary objective in the efforts for
[Knowledge Representation Formalisms and Methods] Semantic creating the thesaurus ten years ago. Therefore it will be pub-
networks; H.3.7 [Digital Libraries] Standards lished 2 now under a Creative Commons (by-nc-sa) license.
The thesaurus community has developed standards for the con-
struction of mono- and multilingual thesauri, but up to now no
Keywords standardized serialization format has emerged 3 . On the other
Thesaurus, SKOS, RDFa, Web Service, REST, Library, Catalog,
hand, the semantic web community has built the “SKOS - Simple
Economics
Knowledge Organization System” [10], which is expected to
achieve the status of a W3C recommendation in early 2009.
1. INTRODUCTION
The “Thesaurus for Economics“ (“Standard Thesaurus 2. RELATED WORK
Wirtschaft“, STW) [6] was developed between 1995 and 1997 A structured method to convert thesauri to SKOS has been devel-
within a publicly funded project for the „Unification of economic oped by van Assem et al. [4]. The authors evaluate the applicabil-
nomenclature“ in Germany. Its main parts cover economics, busi- ity of the SKOS meta model to represent the features of existing
ness economics and sectors of industry, supplemented by related thesauri. Case studies on The Integrated Public Sector Vocabulary
subject areas, geographical and product descriptors. IPSV (UK, government documents), the Common Thesaurus for
Today, the STW is maintained by an editorial board situated at Audiovisual Archives GTAA (NL, TV/radio programs) and the
the German National Library of Economics (ZBW) and used by Medical Subject Headings MeSH (US, biomedical domain) have
scientific institutions, by documentation centers for economics verified the applicability of methodology and model, but also
and business and by GBI-Genios, a provider of business informa- have identified non-standard features it could not handle.
tion databases and one of the co-creators of the thesaurus. Its Later on, other teams (located again at the Vrije Universiteit Am-
5,800 descriptors in German and English are presented within a sterdam) concentrated on the area of cultural heritage. Their ap-
taxonomy of 500 classes to offer an easy navigation along the proach was to use metadata and thesaurus schema mapping,
lines of categories widely accepted in the profession. 17,000 non- metadata mapping and thesaurus alignment to support retrieval
descriptors (mainly in German) direct users to the preferred terms. [16]. They applied this process to six individual collections which
The concepts of the thesaurus are richly interconnected through
15,000 “broader/narrower” and 10,000 “related” links.
1
http://www.econis.eu. The ZBW, as the world’s largest library
for economics, holds about 4 million media units.
Copyright is held by the author/owner(s). 2
LDOW2009, April 20, 2009, Madrid, Spain. http://zbw.eu/stw
3
Recently, a development draft for a XML schema for thesauri
has been published in the context of BS 8723-5 and the upcom-
ing ISO 25964 (http://schemas.bs8723.org/XMLSchema.aspx).
The older Zthes schema (http://zthes.z3950.org) doesn’t seem to
have been widely adopted.
where indexed with six different vocabularies. Thus they could Technically, the conversion process starts on descriptor and tax-
show that unified access to heterogenous collections can be real- onomy files that were, in a custom text format, produced by the
ized using controlled vocabularies and semantic web retrieval thesaurus maintenance application. The files are read by a Perl
tools [11]. script and transformed to internal data structures which are neces-
These activities were partly inspired by undertakings in the cul- sary for term lookup. Making use of RDF::Helper module and the
tural heritage and later in the health information domain in Redland RDF libraries 7 , the SKOS structures are built and
Finland. In a coordinated effort, Finland aims to build a national dumped to an RDF/XML file.
semantic web infrastructure. 4 Open and shared ontologies are This file is enriched with some general information about version-
seen as the main basis for this. The project advocates a more-than- ing, licensing terms and so on, exploiting the fact that SKOS/RDF
syntactic transformation of existing thesauri to SKOS, refining allows the intermixture of properties from arbitrary other ontolo-
and enriching the semantic structures.[7] Another focus is on gies (e.g. dc:publisher, owl:versionInfo or xhtml:license). Also,
public ontology services, such as the ontology server ONKI the SKOS vocabulary and the ZBW vocabulary extensions were
SKOS [18], and on applications such as the HealthFinland portal, added. The generated file serves as a base for all further activities.
destined to provide reliable health information to end-users utiliz-
ing a broad scale of semantic web techniques [15]. 3.1 SKOS Extensions
For the intended use with a broad range of “knowledge organiza-
While interoperability and integration into applications was the
tion systems, such as thesauri, taxonomies, classification schemes
main concern in these approaches, others concentrated on the
and subject heading systems” [10] SKOS provides a very general
accessibility of the data. In an experimental effort, the Library of
framework of classes and properties. Knowledge organization
Congress Subject Headings, one of the largest and most widely
systems “in the wild” therefore often use constructs which have
used terminologies, were presented as interlinked
no direct counterparts in the standard SKOS language. Therefore,
RDFa/RDF/Json on the semantic web. The authors described in
SKOS is designed to make it easy to extend the language con-
detail the process of conversion from MARC to SKOS and the
structs by specialization. Hints on how to do this are given in the
design and deployment as linked data application [14]. After hav-
SKOS documentation.[8] In the case of STW, two extensions
ing abruptly shut off the prototype site, the LC now has an-
have been developed. In our extensions we tried to introduce as
nounced a re-release of the web service. 5
few customizations as possible in order to make usage as easy as
In contrast, the Swedish Union Catalogue, held by the National possible for the unknown user of our data.
Library of Sweden, exposed its data completely on the web as
linked data [9]. Aside from six million bibliographic records, this 3.1.1 Subclassing “skos:Concept”
included also two hundred thousand authority records on authors, The STW is different from many other thesauri in that it includes
titles and subject headings as well as internal and external links. a taxonomy of about 500 classes, sometimes subdivided by up to
For the subject headings a SKOS representation was used. Since five levels. This thesaurus systematics is not used for indexing,
mappings to LCSH had already been created, the Swedish library but as an aid for users to find appropriate indexing or search terms
could easily add links to this information space – “useful for mak- (figure 1).
ing inferences about relations not present in our system of subject
headings from relations present in LCSH” [12].
3. THESAURUS CONVERSION TO SKOS
Following roughly the method given in [4], the conversion of the
STW to SKOS was quite straightforward (apart from two areas
which required custom extensions which are described below):
Main entries build skos:prefLabels, “used for” relations went to
skos:altLabels, and the notes were mapped to skos:scopeNote and
skos:editorialNote. The “broader”, “narrower” and “related” rela-
tions directly mapped to the according SKOS properties. The
built-in multilingual features of SKOS made it easy to handle the
German and English labels connected to the concepts. Other lan-
guages, which were used mainly for the original names of geo-
graphic units and could not be distinguished automatically, were
labeled as “x-other” (according to RFC 4646).
The newly introduced skos:notation was a good fit for the nota-
tion part of the taxonomy entries. The language-dependent label
could have been expressed as skos:prefLabel. But this disregards
the necessity of presenting the notation as part of a meaningful
display label and, even more importantly, as a means for sorting
the entries appropriately. For ease of display through standard
tools, we decided therefore to transform the language-dependent
label to rdfs:label, and to build skos:prefLabel by a concatenation
of skos:notation and this label.
Figure 1. Combined Hierarchy of Taxonomy (in the display rec-
ognizable by the alphanumeric notations) and Descriptors,
4
http://www.seco.tkk.fi/projects/sw20/ as viewed in SKOSEd6 Plugin for Protégé
5
http://lcsh.info/2008/12/19/uncool-uris/; http://id.loc.gov
6 7
http://code.google.com/p/skoseditor/ http://librdf.org
Since the STW includes this taxonomy as well as the descriptors This can be illustrated with an example. Having introduced the
themselves, we wanted to have both in one concept scheme. But distinction between zbwext:Thsys and zbwext:Descriptor de-
we still needed some distinguishability, especially to avoid cum- scribed above, we wanted to verify that a concept of class
bersome phrasing in SPARQL queries. Using skos:Collection for zbwext:Descriptor does not have a concept of class zbwext:Thsys
the taxonomy was not an option, because this class cannot be as skos:narrower. This can be checked by a simple SPARQL
nested to express hierarchies. Therefore we choose to subclass query:
skos:Concept: CONSTRUCT {
zbwext:Descriptor rdfs:subClassOf skos:Concept . [] a :Error;
zbwext:Thsys rdfs:subClassOf skos:Concept . :message 'Descriptor [1] has narrower Thsys [2].';
:implicated ( ?d ?t );
One of the advantages of this approach was that we could use the } WHERE {
well-defined semantics of skos:broader/skos:narrower (and their ?d skos:narrower ?t.
transitive super-properties) to build and query a common hierar- ?d rdf:type zbwext:Descriptor .
chy of taxonomy and thesaurus, and we also could restrict queries ?t rdf:type zbwext:Thsys }
to one of these parts (see example in part 3.2). Thus, Thesaurus maintenance can be facilitated by running a se-
3.1.2 Introducing Structured Notes ries of SPARQL consistency checks on a regular basis to verify
SKOS already provides a broad range of documentation proper- the compliance with custom rules.
ties, but cannot cover every special case. The STW uses, besides
standard “scope note” with explanations about the intended term 4. DEPLOYMENT IN RDFA
usage, also some more formalized notes which link to other de- The deployment of the thesaurus data is meant to serve machines
scriptors. They offer “use this instead of that” hints – e.g., a note as well as humans. The descriptors and the taxonomy are repre-
for the descriptor “Restrictive business practices” says “For re- sented by individual RDFa pages for every concept present in the
strictions on market entry USE Market entry”. To preserve the thesaurus. For human use these pages will be enhanced by a
semantics of this note (and to offer users an easy-to-click naviga- search facility and a taxonomy tree suitable for browsing. Each
tion on the generated web pages), we have introduced a new con- page is linked to other pages through narrower, broader and re-
struct: lated concepts. The URL-rewrite and content negotiation capabili-
zbwext:useInsteadNote rdfs:subPropertyOf skos:note .
ties of the Apache web server are used to do the mapping of
requests, e.g. for http://zbw.eu/stw/descriptor/19664−4 (finan-
Since the range of skos:note deliberately was not restricted to cial crisis):
literals, we can use a blank node as structured value for this prop-
erty. The blank node consists of a string with the literal value for 1) If rdf/xml is requested, a 303 (see other) redirect is performed
the note and a link to the descriptor which should be used instead. to the RDFa Distiller 11 with the address of the current version
For the latter, we did not introduce a special relationship, but English XHTML+RDFa page as a referrer. The user gets the
made use of the general rdfs:seeAlso. The example above can be RDF/XML destilled from this page (as suggested in [3]).
expressed as 2) If html or any other format is requested, a 303 (see
zbwext:useInsteadNote other) redirect is performed to the current version
[ rdf:value "For restrictions on market entry"@en; http://zbw.eu/stw/versions/8.04/descriptor/19664−4/about,
rdfs:seeAlso ] . which is resolved to a language-specific representation by content
negotiation (“Options +MultiViews”). The language-specific
and can be rendered easily to web pages.
pages are also addressable directly.
In a similar fashion, “see-also” notes could be introduced as an-
The web pages are produced from the SKOS file by a Perl script,
other special case of skos:note or skos:scopeNote.
making use of the HTML::Template templating system. All tri-
Nonetheless, applications and readers must be aware of these ples describing a concept are embedded in the page (though some-
custom structures. The use of a new sub-property indicates the times not visible, e.g. skos:altLabels in German on an English
need for further analysis, and SKOS allows it to define the mean- page). For a better support in standard tools we embed the super
ing and structure by the means of skos:definition and/or properties of custom properties also. The RDFa syntax makes this
rdfs:comment. very easy as most attributes are defined as whitespace separated
lists (eg. “typeof= ‘skos:Concept zbwext:Descriptor’”).
3.2 Utilizing SPARQL Queries for Checking
The links to other concepts are dual: the clickable href links to a
Inconsistencies version- und language-specific representation (eg.
The transformation to SKOS allowed us to load the thesaurus into http://zbw.eu/stw/versions/8.04/descriptor/19664−
a SPARQL server (implemented with Joseki 8 , from the Jena pro- 4/about.en.html) in order to make any language choice by a user
ject) and to run all kinds of queries against it. Most useful is the sticky, while the rdf link to the resource is always designated by
possibility to check the data basis for inconsistencies. A series of the generic URI, (eg. http://zbw.eu/stw/descriptor/19664−4).
standard checks can be executed through the validation service of
the Semantic Web Deployment Working Group (SWDWG) 9 . Version information is embedded in the page by additional triples,
Using the code developed for this service 10 as a blueprint, it was with the conceptScheme and the page as a subject. Through an
easy to write custom checks to enforce rules for the STW. overview page all published versions can be navigated and also
downloaded as a RDF/XML zip file. Thus every concept has mul-
tiple pages which describe different statuses of its properties,
while the identity of the concept remains the same. We regard the
8
http://www.joseki.org/ stability of the concept URIs as an essential prerequisite for use in
9
http://www.w3.org/2004/02/skos/validation
10
http://isegserv.itd.rl.ac.uk/cvs-public/~checkout~/skos/drafts/
11
integrity.html?rev=HEAD http://www.w3.org/2007/08/pyRdfa/
indexing and linking to other resources. The “history” of a con- cepts to which interlanguage links from English pages have been
cept and its changes may be looked up by humans. created), while the STW was developed as a German vocabulary
Since the primary objective of the project was to produce web which is enriched with English altLabels only step by step.
pages, embedding additional RDF data into these pages proved an Another restriction resulted from the unforeseen impact of the
easy way for “semantic enrichment”. The main effort there was to adoption of “USED FOR” relations in the STW. In some places,
figure out the adequate web server configuration settings. particularly in the products branch, it uses “upward posting” – e.g.
“pineapple”, “avocado”, “cola nut” etc. are used for and become
skos:altLabels for the descriptor “”. This is per-
5. LINKING TO OTHER RESOURCES fectly compliant with ISO 2788 [1] and SKOS. Although not rec-
Providing links to other resources inside and outside the library ommended there, it works well in the closed world of a library
was a main design target of the new STW web presentation. The collection and effectively supports both indexing and retrieval.
first and most natural aim for this was ECONIS, our own library However, it renders the use of skos:altLabels for matching con-
catalog. The concepts of the thesaurus are imported as an author- cepts in an open world awkward. In DBpedia, “pineapple”, “avo-
ity file into the catalog and have been used for indexing the books cado” and so on are concepts of their own, and it would be
and other media of the library for many years. misleading to assume a match with the concept “”.
This issue could be solved only by intellectual discrimination
5.1 Library catalog within the thesaurus maintenance system.
URLs which execute a search in the library system can be easily
constructed, with ID values derived from the concept URI or with The high amount of unsucessful matches is not only the result of
German or English literal terms as search terms. We decided to simple derivations in the compared strings. There is also a (not
use the skos:prefLabel, in the user’s favored language, because yet quantified but significant) number of economics concepts
the search term displayed on the result page is much more infor- which do not exist as entries in DBpedia. Examples of these are
mative than an ID. “agricultural price”, “annual audit” or “youth unemployment”.
While this is disappointing for the task of linking the data sets, it
The links embedded in the page should not only be navigable by
underlines the undeniable importance of technical terminologies
human users but also by semantic web crawlers (even though up
as linking hubs for specialized communities.
to now the result pages of the library system lack semantic infor-
mation). Since the skos:subject/isSubjectOf properties ceased to
exist, there was no easily fitting property for this purpose. Al- 6. BUILDING A THESAURUS BASED
though it would be questionable to define an inversePropertyOf, TERMINOLOGY WEB SERVICE
eg. dc:subject, given the large and hardly useful amount of such It was recognized long ago that thesaurus resources could be util-
inverse relations, in our opinion it is clearly desirable to be able to ized to support retrieval. The paradigm of service orientation in
link to collections of resources which are indexed with a given the software universe suggests implementing a terminology ser-
concept. As a standard manner has not been defined yet, we intro- vice as a web service. A RPC-style, SOAP-based SKOS API for
duced a custom property such a service has been defined early on [2]. It has some imple-
zbwext:indexedItem rdfs:subPropertyOf rdfs:seeAlso mentations (e.g. the FACET browser [17]) but appears to have
seen no further development. In contrast to these beginnings our
to express the special meaning of the link. approach is based on a REST architecture. It tries to expose
meaningful and useful thesaurus-related resources on the web,
5.2 DBpedia available through simple HTTP GET requests, using resource
While we could generate these links unhesitatingly, since we
names (like “concepts”) rather than verbs (like “getConcepts”).
knew in advance that each would hit, this is not true in the open
world of other datasets. To connect to the Linked Data Cloud, A beta version of the STW web service 12 has defined three basic
DBpedia[5] was chosen as the main linking hub. An alignment resources:
with DBpedia resources also establishes the chance to offer useful concepts – searches and returns concept URIs for a given search
information from Wikipedia to the users of the STW. term or query (for ease of handling in an application, eg. for dis-
The matching was done in a two-step process: First, match all play in a list, accompanied by skos:prefLabels). By default,
concept labels (preferred and alternate, German and English) search is restricted to the type of zbwext:Descriptor and carried
against the DBpedia labels and redirects (as lowercased strings). out on skos:prefLabel and skos:altLabel
Second, evaluate these matches and build relations where the narrower – return the skos:narrower concept URIs for a given
match can be trusted – if at least one prefLabel matches. If both concept URI. (broader and related to be defined in the same way
English and German prefLabels match, we assume a to cover the other semantic relations)
skos:exactMatch, otherwise a skos:closeMatch.
labels – returns (by default: all) labels for a given concept (ac-
raw matches en de total descriptors 5766 companied by the concept URI and its skos:prefLabel)
prefLabel 3018 1804 sucessful match 3042
These are the basic building blocks. They could be useful in
altLabel 1081 1822 both prefLabel 1242
themselves, for example to enhance a search within a dataset in-
total 4099 3626 single prefLabel 1295 dexed by the concepts of a given concept scheme. If the search
pref- and altLabel 505 term can be mapped to a concept URI by the search application, it
unsucessful match 2724 can include narrower concept URI via a lookup of the “narrower”
no match at all 1903 service for this concept URI.
no prefLabel match 552
However, mostly the user input has to be mapped to concepts, and
prefLabels differ 269 then resources related to these concepts are demanded. So we
As the results of the raw matches show, the different language
emphases of the datasets matter: DBpedia is built around the Eng-
lish version of Wikipedia (and includes only those German con- 12
http://zbw.eu/beta/stw-ws
define resources which represent combinations of these basic 5. Auer, S.; Bizer, C.; Lehmann, J.; Kobilarov, G.; Cyganiak, R.;
building blocks, e.g. Ives, Z. DBpedia: A Nucleus for a Web of Open Data.
synonyms – returns synonyms for a given search term or query Aberer et al. (Eds.): The Semantic Web, 6th International
by chaining concepts and labels. To give an example: Semantic Web Conference, 2nd Asian Semantic Web Con-
http://zbw.eu/beta/stw−ws/synonyms?query=free+trade+zone ference, Springer (2007).
returns the terms “Export processing zone”, “Foreign trade zone”, 6. Gastmeyer, M. and (Red.). Standard-Thesaurus Wirtschaft.
“Foreign-trade zone”, and optionally equivalent German terms. Deutsche Zentralbibliothek für Wirtschaftswissenschaften,
The “packaging” of basic services is not only a matter of conven- Kiel, 1998.
ience, but much more one of performance. Since the services must 7. Hyvönen, E., Viljanen, K., Mäkelä, E., et al. Elements of a
be well suited for use in search engines for economics resources, National Semantic Web Infrastructure - Case Study Finland
it is absolutely critical to get all necessary details required for on the Semantic Web (Invited paper). Proceedings of the
query expansion in one and only one round-trip. First International Semantic Computing Conference (IEEE
Currently the web service is already the foundation of the search ICSC 2007), Irvine, California, (2007).
interface for the thesaurus itself which implements a Google Sug- 8. Isaac, A. and Summers, E. SKOS Simple Knowledge Organi-
gest-like, AJAX-based incremental search. This gives us an op- zation System Primer. W3C Working Draft, 2008.
portunity to explore the real-time characteristics of the service and http://www.w3.org/TR/skos-primer/.
the possibilities and impacts of standard web-caching mechanisms
9. Malmsten, M. Making a Library Catalogue Part of the Seman-
on performance. An integration into a new version of the library's
tic Web. Proc. Int’l Conf. on Dublin Core and Metadata Ap-
document server retrieval interface based on Lucene could give
plications, (2008).
some feedback about the usefulness of the service when applied to
document retrieval and the scalability of the implementation un- 10. Miles, A. and Bechhofer, S. SKOS Simple Knowledge Or-
der load. ganization System Reference. W3C Working Draft, 2008.
http://www.w3.org/TR/skos-reference/.
To explore other kinds of queries, we plan to expose an experi-
mental public SPARQL endpoint based (as the service described 11. Schreiber, G., Amin, A., Aroyo, L., et al. Semantic annotation
above) on a Joseki server. and search of cultural-heritage collections: The Multime-
diaN E-Culture demonstrator. Web Semantics: Science, Ser-
7. CONCLUSION vices and Agents on the World Wide Web 6, 4 (2008), 243-
249.
Our choice of SKOS and RDF as a publication and exchange
format for the “Thesaurus for Economics” has proved successful. 12. Söderbäck, A. and Malmsten, M. LIBRIS - Linked Library
We could build on this format for generating a richly interlinked Data. Nodalities blog. From Semantic Web to Web of Data,
website. The web pages in turn embed the semantic information 2009. http://blogs.talis.com/nodalities/2009/01/libris-linked-
as RDFa and are connected to the Linked Data Cloud on the web. library-data.php.
Further services such as consistency checking and terminological 13. Summers, E. following your nose to the web of data. inkdroid
web services could be implemented on the same basis. Blog Archive, 2008.
A logical future step would be the presentation of links to other http://inkdroid.org/journal/2008/01/04/following-your-nose-
terminologies and classifications (eg. the widely used classifica- to-the-web-of-data/.
tion of the “Journal of Economic Literature”). Especially if these 14. Summers, E., Isaac, A., Redding, C., and Krech, D. LCSH,
terminologies and classifications are on the semantic web them- SKOS and Linked Data. 0805.2855, 2008.
selves, new services, such as “translation” services for search http://arxiv.org/abs/0805.2855.
queries, are possible. While thesauri provide no strict ontologies,
and therefore no basis for automated reasoning, as carefully 15. Suominen, O., Hyvönen, E., Viljanen, K., and Hukka, E.
crafted terminologies they can be highly useful for inter-linking HealthFinland – A National Publication System for Seman-
knowledge resources in libraries and beyond. tic Health Information. Semantic Computing Research
Group, Helsinki University of Technology and University of
Helsinki, 2008.
8. REFERENCES http://www.seco.tkk.fi/publications/2008/suominen-et-al-
1. ISO 2788:1986 Documentation - Guidelines for the establish- healthfinland-swc-2008.pdf.
ment and development of monolingual thesauri. 1986.
16. Tordai, A., Omelayenko, B., and Schreiber, G. Semantic Ex-
2. SKOS API - SWAD-Europe Thesaurus Activity. 2004. cavation of the City of Books. SAAKM, CEUR-WS.org
http://www.w3.org/2001/sw/Europe/reports/thes/skosapi.htm (2007).
l.
17. Tudhope, D. and Binding, C. Towards terminology services:
3. How to add RDF information to a page using RDFa? W3C Experiences with a pilot web service thesaurus browser.
Q&A Weblog, 2008. http://dcpapers.dublincore.org/ojs/pubs/article/viewArticle/8
http://www.w3.org/QA/2008/05/using_rdfa_to_add_informa 31.
tion.html.
18. Tuominen, J., Frosterus, M., Viljanen, K., and Hyvönen, E.
4. van Assem, M., Malaisé, V., Miles, A., and Schreiber, G. A ONKI-SKOS ― Publishing and Utilizing Thesauri in the
Method to Convert Thesauri to SKOS. In The Semantic Semantic Web. AI and Machine Consciousness - Proceed-
Web: Research and Applications. 2006, 95-109. ings of the 13th Finnish Artificial Intelligence Conference
STeP 2008, (2008).