A presentation of the project Culturegraph Authorities at the PDR workshop in Berlin on March 6th, 2013 (http://pdr.bbaw.de/veranstaltungen/pdr-workshop-2013). Other topics were the use of the GND-Ontology (http://d-nb.info/standards/elementset/gnd) and AgRelOn (http://d-nb.info/standards/elementset/agrelon) to prepare and describe data about persons for use in a linked data context. The talk and the presentation were co-prepared by Markus Geipel, Christoph Böhme and myself.
Interoperability and Standards : Thoughts on (bibliographic) Data Exchange af...Lars G. Svensson
Libraries are moving towards the (Semantic) Web as the main platform for data exchange. In order to keep the data interoperable, they need to consider with whom they want to exchange information and how to create network nodes that others can connect to. The working hypothesis is that the might be achieved by more heavy use of the LRM Work entity. Another point is the need to keep cataloguing codes, data models and data exchange formats as separate as possible.
IFLA LIDASIG Open Session 2017: Introduction to Linked DataLars G. Svensson
At the IFLA Linked Data Special Interest Group open session in Wroclaw we briefly introduced the mission of the SIG and then went on to a brief introduction to what linked data is and why that topic is important to libraries.
The presentation was held jointly by Astrid Verheusen (general introduction to the SIG) and Lars G. Svensson (introduction to Linked Data)
Towards an Authoritative Global Data Infrastructure: Connecting Libraries wit...Lars G. Svensson
With Fake News and Alternative Facts being part of the everyday agenda, the notion of trustworthy data is getting increasingly important. Government agencies have an important role to play here as suppliers of authoritative information. In this presentation, held in the Edinburgh University Library on February 27, 2017, I started out outlining what data we publish in the German National Library and then go on to explore how this information can be interesting to other government agencies and what information they publish can be interesting for libraries.
Relations matter: Maintaining and Publishing Links in Library Metadata Lars G. Svensson
In order for library (meta-)data to be published as linked data, it has to contain links. At the OCLC EMEA Meeting in Berlin on February 21-22, 2017, I used those slides to present how we create, ingest, store and export links to and from the metadata we create in the German National Library
In http, media types are often used in content negotiation, but those can often only say something about the format of the data, not about the semantics used within the data. This is particularly an issue with data in RDF where the same resource can be described in several different ways using different RDF vocabularies (e. g. DublinCore, foaf or schema,org). This presentation and the accompanying position paper (https://www.w3.org/2016/11/sdsvoc/SDSVoc16_paper_14) presented at SDSVoc in Amsterdam, I suggest a new http header to resolve this problem.
When deciding on how to describe cultural heritage resources in common exchange formats (e. g. MARC 21, RDF or XML), publishing organisations need to align their content standards with wide-spread, broadly adopted data standards in order to make information exchange as effective as possible.
This presentation from the IFLA Committee on Standards session in Cape Town on August 19, 2015 (2015-08-19) makes that case. There is also an accompanying paper in the IFLA library at http://library.ifla.org/id/eprint/1194
What do we need to consider when we map knowledge organisation systems to each other? A EDUG (European Dewey User Group) workshop in Naples looked at this question with particular attention to the Dewey Decimal Classification (DDC). This presentation by me and Tina Mengel discusses the experiences we drew from the CrissCross project (using the German subject headings to access information classified using the DDC), and then goes on to analyse what needs to be done in order to publish those mappings, both in traditional library formats and also as linked data.
Folien für das Seminar "Bibliotheksdatenpublikation und Linked Data" in der Österreichischen Nationalbibliothek am 11. September 2014 (Brainpool K 04/2014). Der erste Teil behandelt die Repräsentation von Bibliotheksdaten als RDF (Grundlagen, Modelle), der zweite Teil ist eine Einführung in das Framework Metafacture; hier schulde ich Christoph Böhme von der DNB großen Dank, da ich auf seine Folien zurückgreifen konnte.
Interoperability and Standards : Thoughts on (bibliographic) Data Exchange af...Lars G. Svensson
Libraries are moving towards the (Semantic) Web as the main platform for data exchange. In order to keep the data interoperable, they need to consider with whom they want to exchange information and how to create network nodes that others can connect to. The working hypothesis is that the might be achieved by more heavy use of the LRM Work entity. Another point is the need to keep cataloguing codes, data models and data exchange formats as separate as possible.
IFLA LIDASIG Open Session 2017: Introduction to Linked DataLars G. Svensson
At the IFLA Linked Data Special Interest Group open session in Wroclaw we briefly introduced the mission of the SIG and then went on to a brief introduction to what linked data is and why that topic is important to libraries.
The presentation was held jointly by Astrid Verheusen (general introduction to the SIG) and Lars G. Svensson (introduction to Linked Data)
Towards an Authoritative Global Data Infrastructure: Connecting Libraries wit...Lars G. Svensson
With Fake News and Alternative Facts being part of the everyday agenda, the notion of trustworthy data is getting increasingly important. Government agencies have an important role to play here as suppliers of authoritative information. In this presentation, held in the Edinburgh University Library on February 27, 2017, I started out outlining what data we publish in the German National Library and then go on to explore how this information can be interesting to other government agencies and what information they publish can be interesting for libraries.
Relations matter: Maintaining and Publishing Links in Library Metadata Lars G. Svensson
In order for library (meta-)data to be published as linked data, it has to contain links. At the OCLC EMEA Meeting in Berlin on February 21-22, 2017, I used those slides to present how we create, ingest, store and export links to and from the metadata we create in the German National Library
In http, media types are often used in content negotiation, but those can often only say something about the format of the data, not about the semantics used within the data. This is particularly an issue with data in RDF where the same resource can be described in several different ways using different RDF vocabularies (e. g. DublinCore, foaf or schema,org). This presentation and the accompanying position paper (https://www.w3.org/2016/11/sdsvoc/SDSVoc16_paper_14) presented at SDSVoc in Amsterdam, I suggest a new http header to resolve this problem.
When deciding on how to describe cultural heritage resources in common exchange formats (e. g. MARC 21, RDF or XML), publishing organisations need to align their content standards with wide-spread, broadly adopted data standards in order to make information exchange as effective as possible.
This presentation from the IFLA Committee on Standards session in Cape Town on August 19, 2015 (2015-08-19) makes that case. There is also an accompanying paper in the IFLA library at http://library.ifla.org/id/eprint/1194
What do we need to consider when we map knowledge organisation systems to each other? A EDUG (European Dewey User Group) workshop in Naples looked at this question with particular attention to the Dewey Decimal Classification (DDC). This presentation by me and Tina Mengel discusses the experiences we drew from the CrissCross project (using the German subject headings to access information classified using the DDC), and then goes on to analyse what needs to be done in order to publish those mappings, both in traditional library formats and also as linked data.
Folien für das Seminar "Bibliotheksdatenpublikation und Linked Data" in der Österreichischen Nationalbibliothek am 11. September 2014 (Brainpool K 04/2014). Der erste Teil behandelt die Repräsentation von Bibliotheksdaten als RDF (Grundlagen, Modelle), der zweite Teil ist eine Einführung in das Framework Metafacture; hier schulde ich Christoph Böhme von der DNB großen Dank, da ich auf seine Folien zurückgreifen konnte.
Jusqu'où l'interopérabilité est-elle nécessaire?Lars G. Svensson
How much interoperability does library metadata need in order to be useful (within and outside of libraries). There is no clear answer to that question and the first attempt (42) is a hint that we might have to pose the question differently.
Those slides were accompanying my keynote speech at the Journées ABES in Montpellier on May 20, 2014.
When the web moves from strings to things libraries can leverage on their authority data to improve the navigation in their catalogues and thus to help improve the user experience.
In this presentation held at a workshop organised by the URBS library community in Rome on November 18, 2013, I try to give theoretical and practical examples of how this can be implemented.
MARC is dying, some people say, but what comes instead? The Library of Congress has initiated the Bibliographic Framework Transition Initiative (BIBFRAME) that develop a transition path from traditional, records-based bibliographic descriptions to a model built on linked data principles. Deutsche Nationalbibliothek (German National Library, DNB) is among the early implementers. This presentation held at SWIB 2013 in Hamburg gives an overview of the initiative, the activities performed by the DNB so far and what the future might look like.
A presentation about the linked data activities in the German National Library accompanying my lunch talk in the National Library of New Zealand on August 13, 2013.
Linked data in the German National Library at the OCLC IFLA round table 2013Lars G. Svensson
A presentation about the current state of the linked data activities in the German National Library held at the OCLC Linked Data Round table during the WLIC 2013 in Singapore
A presentation at the workshop "Rich and loonely or poor and popular?" at the Dublin Core conference in Lisbon on September 4th, 2013. The main hypothesis is that when publishing (linked) data, the main criteria should not be richness and poorness, but suitability for purpose, granularity and adherence to agreed-on models.
The German National Library (Deutsche Nationalbibliothek, DNB) is currently changing its business model to publish all data in all formats under a CC0 license. This is a presentation of the current state I made on the Open Data on the Web conference organised by the W3C, the Open Data Institute and the Open Knowledge Foundation in London on April 24, 2013. There is an accompanying paper available at http://www.w3.org/2013/04/odw/odw13_submission_57.pdf.
Features for the future: What Might the Dewey Ecosystem Look Like in Ten Years?Lars G. Svensson
In the last ten years there has been much change in the Dewey ecosystem. One driver has been the European Dewey User Group (EDUG). This presentation, which I held at the EDUG meeting in the National Library of Norway, Oslo in April 2013, tries a look into the crystal ball in order to find out what the Dewey ecosystem looks like in ten years' time. The main points:
1) Dewey has changed some internals, including licensing, versioning and use of facets
2) DDC interoperates well with other knowledges information systems and we have well-established crosswalks in place in order to improve information retrieval
3) The integration of classification data into end-user interfaces is mainstream
If this becomes true depends on the people who use the system! If they want it to happen, it will.
Linked data activities in the Deutsche NationalbibliothekLars G. Svensson
This presentation accompanied a lightning talk at the IFLA Semantic Web Special Interest Group's session at the World Library and Information Conference in Helsinki 2012
The Deutsche Nationalbibliografie as linked open dataLars G. Svensson
This presentation was held at the World Library and Information Conference 2012 in Helsinki. It is a companion to the position paper by Jürgen Kett et al. (http://nbn-resolving.de/urn:nbn:de:101-2012052306) which discusses the requirements on a national bibliography in the 21st century. The paper argues that a national bibliography needs to integrate itself into the World Wide Web, since this is where information exchange takes place today. This should be made using linked data technologies and the data should be published under an open license. As a case study, we present the work done at the Deutsche Nationalbibliothek, where circa 70% of the database is already published as linked open data.
Jusqu'où l'interopérabilité est-elle nécessaire?Lars G. Svensson
How much interoperability does library metadata need in order to be useful (within and outside of libraries). There is no clear answer to that question and the first attempt (42) is a hint that we might have to pose the question differently.
Those slides were accompanying my keynote speech at the Journées ABES in Montpellier on May 20, 2014.
When the web moves from strings to things libraries can leverage on their authority data to improve the navigation in their catalogues and thus to help improve the user experience.
In this presentation held at a workshop organised by the URBS library community in Rome on November 18, 2013, I try to give theoretical and practical examples of how this can be implemented.
MARC is dying, some people say, but what comes instead? The Library of Congress has initiated the Bibliographic Framework Transition Initiative (BIBFRAME) that develop a transition path from traditional, records-based bibliographic descriptions to a model built on linked data principles. Deutsche Nationalbibliothek (German National Library, DNB) is among the early implementers. This presentation held at SWIB 2013 in Hamburg gives an overview of the initiative, the activities performed by the DNB so far and what the future might look like.
A presentation about the linked data activities in the German National Library accompanying my lunch talk in the National Library of New Zealand on August 13, 2013.
Linked data in the German National Library at the OCLC IFLA round table 2013Lars G. Svensson
A presentation about the current state of the linked data activities in the German National Library held at the OCLC Linked Data Round table during the WLIC 2013 in Singapore
A presentation at the workshop "Rich and loonely or poor and popular?" at the Dublin Core conference in Lisbon on September 4th, 2013. The main hypothesis is that when publishing (linked) data, the main criteria should not be richness and poorness, but suitability for purpose, granularity and adherence to agreed-on models.
The German National Library (Deutsche Nationalbibliothek, DNB) is currently changing its business model to publish all data in all formats under a CC0 license. This is a presentation of the current state I made on the Open Data on the Web conference organised by the W3C, the Open Data Institute and the Open Knowledge Foundation in London on April 24, 2013. There is an accompanying paper available at http://www.w3.org/2013/04/odw/odw13_submission_57.pdf.
Features for the future: What Might the Dewey Ecosystem Look Like in Ten Years?Lars G. Svensson
In the last ten years there has been much change in the Dewey ecosystem. One driver has been the European Dewey User Group (EDUG). This presentation, which I held at the EDUG meeting in the National Library of Norway, Oslo in April 2013, tries a look into the crystal ball in order to find out what the Dewey ecosystem looks like in ten years' time. The main points:
1) Dewey has changed some internals, including licensing, versioning and use of facets
2) DDC interoperates well with other knowledges information systems and we have well-established crosswalks in place in order to improve information retrieval
3) The integration of classification data into end-user interfaces is mainstream
If this becomes true depends on the people who use the system! If they want it to happen, it will.
Linked data activities in the Deutsche NationalbibliothekLars G. Svensson
This presentation accompanied a lightning talk at the IFLA Semantic Web Special Interest Group's session at the World Library and Information Conference in Helsinki 2012
The Deutsche Nationalbibliografie as linked open dataLars G. Svensson
This presentation was held at the World Library and Information Conference 2012 in Helsinki. It is a companion to the position paper by Jürgen Kett et al. (http://nbn-resolving.de/urn:nbn:de:101-2012052306) which discusses the requirements on a national bibliography in the 21st century. The paper argues that a national bibliography needs to integrate itself into the World Wide Web, since this is where information exchange takes place today. This should be made using linked data technologies and the data should be published under an open license. As a case study, we present the work done at the Deutsche Nationalbibliothek, where circa 70% of the database is already published as linked open data.
The Deutsche Nationalbibliografie as linked open data
Aggregierte Präsentation ausgehend von der GND: Culturegraph Authorities, GND-Ontologie und AgRelOn
1. Lars G. Svensson, Markus Geipel, Christoph Böhme
Aggregierte Präsentation dezentral
gehaltener Personendaten: Perspektiven
und semantische Modellierung
ausgehend von der Gemeinsamen
Normdatei (GND)
1 | 23 | Culturegraph Authorities | 06. März 2013
2. Personen nehmen in der historischen
Photo von ernop (CC BY): http://www.flickr.com/photos/kouchi/1165602380/
Forschung einen zentralen Platz ein
2 | 23 | Culturegraph Authorities | 06. März 2013
3. Jeder, der Personendaten verwendet,
hat wahrscheinlich von der GND gehört
3 | 23 | Culturegraph Authorities | 06. März 2013
4. Viele Dienste referenzieren die GND, die
GND verlinkt aber nur auf wenige Dienste
Duplikate ?
GND
Fehlende
links
4 | 23 | Culturegraph Authorities | 06. März 2013
5. Wir brauchen ein System, das alle
verknüpften Daten und Datenquellen zeigt
Versteckte
Beziehungen
Duplikate
Fehlende ?
Links
zurück
GND
Fehlende
links
5 | 23 | Culturegraph Authorities | 06. März 2013
6. Eine Lösung soll Culturegraph Authorities
liefern
Versteckte
Beziehungen
culturegraph.org
Duplikate
Fehlende ?
!
!
Links
zurück
GND
Fehlende
links
6 | 23 | Culturegraph Authorities | 06. März 2013
7. Ziel von Culturegraph Authorities ist,
Normdaten als Einstiegspunkt anzubieten
BVB
BSZ
hbz
HeBIS
GBV
GND
Andere
Institutionen
…
7 | 23 | Culturegraph Authorities | 06. März 2013
8. Dreh- und Angelpunkt ist die
Gemeinsame Normdatei (GND)
Geister Schlagwörter
Personen Götter
Projekte Familien
Produkte Körperschaften
Geographica Werke
8 | 23 | Culturegraph Authorities | 06. März 2013
9. Culturegraph Authorities baut auf der
Plattform Culturegraph auf
120 Millionen Datensätze verbunden durch fünf Algorithmen
culturegraph.org
9 | 23 | Culturegraph Authorities | 06. März 2013
10. Die Arbeit mit den Daten fokussiert auf
drei Aufgaben:
1. Herstellen von Verbindungen
2. Analyse von Verbindungen
3. Verwendung von Verbindungen
10 | 23 | Culturegraph Authorities | 06. März 2013
11. Wir sind auf der Suche nach neuen
Partnern, um die Plattform auszubauen
hbz HeBIS
Hier könnte
SWB
Culturegraph Ihr Name
stehen!
Historische
DNB Kommission
der BAW
11 | 23 | Culturegraph Authorities | 06. März 2013
12. Verknüpfungen entstehen über statistische
Verfahren oder über Konkordanzen
<mx:datafield tag="153" ind2=" " ind1=" ">
#FORMAT: BEACON
<mx:subfield code="a">220.531</mx:subfield>
@prefix void: <http://rdfs.org/ns/void#> . <mx:subfield code="y">1</mx:subfield>
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> . <mx:subfield code="a">7</mx:subfield>
@prefix
@prefix
rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
owl: <http://www.w3.org/2002/07/owl#> .
<mx:subfield code="e">220.531</mx:subfield>
<mx:subfield code="j">Paraphrasen</mx:subfield>
#PREFIX: http://d-nb.info/gnd/ #TARGET:
http://cgi-host.uni-
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> . <mx:subfield code="9">ess=ren</mx:subfield>
@prefix dcterms: <http://purl.org/dc/terms/> . </mx:datafield>
@prefix foaf: <http://xmlns.com/foaf/0.1/> . </mx:record>
marburg.de/~omgesa/gs/xs1.php?
f1=pnd&s1={ID}
#FEED: http://www.online.uni-
marburg.de/fpmr/pnd.txt #CONTACT: Jörg
Witzel joerg.witzel@staff.uni-marburg.de
#INSTITUTION: Forschungsstelle für
Personalschriften
#NAME: Gesamtkatalog deutschsprachiger
Leichenpredigten (GESA)
#MESSAGE: Leichenpredigten 1550-1800
(GESA) #TIMESTAMP: 2012-08-09
100000193|3
100006353|9
<owl:AnnotationProperty rdf:about="http://d-nb.info/standards/elementset/gnd#marc21equivalent"> 100009271|3
<rdfs:label xml:lang="de">MARC 21 Entsprechung</rdfs:label>
<rdfs:label xml:lang="en">MARC 21 equivalent</rdfs:label>
<rdfs:comment xml:lang="de">Gibt die MARC 21 Entsprechung des Elements im GND-MARC-Format wieder.</rdfs:comment>
100017061|9
<rdfs:comment xml:lang="en">Expresses the MARC 21 equivalent for this element in the GND MARC format.</rdfs:comment>
</owl:AnnotationProperty> 100017495|1
100018335|8
100019552|5
12 | 23 | Culturegraph Authorities | 06. März 2013 100021549|3
100029752|199
13. Bisher kommen die meisten Daten aus
den Bibliotheksverbünden
– Schlagwörter – Personenreferenzen
– (Deutsche – (Deutsche Nationalbibliografie)
Bibliotheksverbünde)
1. Deutschland 1. Goethe, Johann Wolfgang
2. Geschichte 2. Kant, Immanuel
3. USA 3. Nietzsche, Friedrich
4. Literatur 4. Jesus Christus
5. Deutsch 5. Heidegger, Martin
6. Frankreich 6. Hegel, Georg Wilhelm Friedrich
7. Zeitschrift 7. Mozart, Wolfgang Amadeus
13 | 23 | Culturegraph Authorities | 06. März 2013
14. Eine weitere
Datenquelle ist die
Wikipedia
Johanna Spyri
14 | 23 | Culturegraph Authorities | 06. März 2013
15. Zur Strukturierung der Daten bieten wir
die GND-Ontologie und AgRelOn
Wolfgang
gnd: Amadeus
118584596 Mozart
gndont:
Differentiated rdf:type agrelon:hasChild
Person
gnd: Franz Xaver
119176653 Wolfgang
Mozart
http://d-nb.info/standards/elementset/gnd
http://d-nb.info/standards/elementset/agrelon
15 | 23 | Culturegraph Authorities | 06. März 2013
16. Die GND-Ontologie beschreibt die Daten
der GND
– entitätenbasierte Repräsentation der GND-Daten
- Personen, Familien, Körperschaften, Kongresse und
Veranstaltungen, Geografika, Schlagwörter, Werke
– eine Ontologie zur Beschreibung der GND entwickeln
- Voraussetzung:
1. entitätenbasierter GND-Erfassungsleitfaden
2. Beschreibung der verwendeten Felder und Codes
– Bereitstellung kontrollierter Vokabulare in RDF/SKOS
- Sprachencodes, Ländercodes, GND-Sachgruppen,
Geschlechtsangaben etc.
16 | 23 | Culturegraph Authorities | 06. März 2013
17. AgRelOn ist speziell auf die Darstellung
von Personenbeziehungen ausgerichtet
17 | 23 | Culturegraph Authorities | 06. März 2013
18. Es gibt kleine Schnittmengen zwischen
GND-Ontologie und AgrRelOn
… … …
bete Beteiligte Tf; Tb; Ts; Tu
beza Bekanntschaft mit Tp
bezb Beziehung beruflich Tp
bezf Beziehung familiär Tp
bilh Bildhauer Tg; Tu
… … …
malr Maler Tu
mitg Mitglied Tp (008 pxl;pxg;pif)
musi Musiker Tb; Tf
… … …
18 | 23 | Culturegraph Authorities | 06. März 2013
19. 19 | 23
zu navigieren?
| Culturegraph Authorities | 06. März 2013
So können Normdaten helfen, im WWW
Photo by Nico Kaiser (CC-BY): http://www.flickr.com/photos/nicokaiser/4667377944/sizes/z/in/photostream/
20. Es gibt kaum Wege von der GND zu
anderen Diensten, die brauchen wir aber!
Versteckte
Beziehungen
Duplikate
Fehlende ?
Links
zurück
GND
Fehlende
links
20 | 23 | Culturegraph Authorities | 06. März 2013
21. Culturegraph Authorities soll eine
zentrale Lösung dafür bieten
Versteckte
Beziehungen
culturegraph.org
Duplikate
Fehlende ?
!
!
Links
zurück
GND
Fehlende
links
21 | 23 | Culturegraph Authorities | 06. März 2013
22. Wir laden dazu ein, eine gemeinsame
Plattform aufzubauen
Photo by sekihan (CC BY-NC): http://www.flickr.com/photos/sekihan/5126922008/
Culturegraph
signup
22 | 23 | Culturegraph Authorities | 06. März 2013