lodlam summit session browsable linked data
0 likes•300 views
This document discusses developing a distributed network of digital heritage information in the Netherlands. It proposes taking a resource-centric linked data approach, implementing linked data principles in data sources, building a knowledge graph, and creating a registry to link organizations, datasets, and resources. This would allow for federated querying across distributed data sources and improved discovery of digital heritage information.
![Design issues...
A tiny example...suppose a resource exists:
<http://data.museum1.nl/object1>
a nde:painting ;
dcterms:subject <http://knowledgegraph.nde.nl/term1> .
For ‘browsable Linked Data’ you would like to add the inverse relation:
<http://knowledgegraph.nde.nl/term1>
a skos:Concept ;
skos:prefLabel "Flowers" ;
skos:isSubjectOf <http://data.museum1.nl/object1> .
TBL about ‘browsable linked data’ [1]:
In practice, when data is stored in two documents, this means that any RDF statements which relate
things in the two files must be repeated in each.
[1]: https://www.w3.org/DesignIssues/LinkedData.html](https://image.slidesharecdn.com/20170627lodlamsummitsessionadistributednetworkofdigitalheritageinformation-170628091005/85/lodlam-summit-session-browsable-linked-data-11-320.jpg)
![Possible approaches to make it work
• Just implement schema.org and let search engines do the work…
• Aggregate all the related Linked Data resources:
- collect datadumps and build a large triplestore
- impressive work f.e. with LOD Laundromat
• Use federated querying:
- doesn’t work very well using regular triplestores
- implement Linked Data Fragments, but what if we have 1500 nodes?
- use logic for predicting most relevant endpoint [1]?
[1]: Routing Memento Requests Using Binary Classifiers](https://image.slidesharecdn.com/20170627lodlamsummitsessionadistributednetworkofdigitalheritageinformation-170628091005/85/lodlam-summit-session-browsable-linked-data-12-320.jpg)
![TBL on Decentralization:
This is a principle of the design of distributed systems, including societies. It points out that
any single common point which is involved in any operation trends to limit the way the
system scales, and produce a single point of complete failure.
Interesting work:
● Interoperability and FAIRness [1]
● IPFS and Linked Open Data [2]
● https://discuss.ipfs.io/t/whos-using-ipfs-in-libraries-archives-and-museums/130
[1]: Interoperability and FAIRness through a novel combination of Web technologies - Mark D. Wilkinson, Ruben Verborgh et al. - April 2017 -
https://peerj.com/articles/cs-110/
[2]: Sharing Linked Open Data over Peer-to-Peer Distributed File Systems: The Case of IPFS - Miguel-Angel Sicilia, Salvador Sánchez-
Alonso - November 2016 - DOI: 10.1007/978-3-319-49157-8_1
A distributed network of digital heritage information??](https://image.slidesharecdn.com/20170627lodlamsummitsessionadistributednetworkofdigitalheritageinformation-170628091005/85/lodlam-summit-session-browsable-linked-data-18-320.jpg)
lodlam summit session browsable linked data
- 1. A distributed network of digital heritage information Enno Meijers
- 2. Netwerk Digital Erfgoed (NDE) aims at increasing the social value of the (digital) heritage information maintained by libraries, archives, museums and other cultural heritage institutions. Long term cooperation between the government and the institutions on national, regional and local level. It’s about information and people!
- 3. Three layered approach for improving the sustainability, the usability and the visibility of digital heritage information.
- 5. Typical aggregation process collect clean enrich harmonize select index OAI-PMH based => data is copied isolated improvements => datasource stays unchanged one size fits all => possible loss of details present aimed at presentation => possible inefficient use of data presentation at portal level => no feedback to the datasource Based on a repository-centric approach...
- 6. And then build networks of aggregators...
- 7. Europeana Digital Public Library of America And it is a lot of work...
- 9. Sören Auer (Fraunhofer): Enterprise Knowledge Graphs - 2016
- 10. Developing a resource-centric approach First step: Implement Linked Data principles in the datasources • use formal (sustainable!) URIs that describe the resources - specify a URI strategy for the Digital Heritage Network - implement persistent URIs (handles) • provide resource URIs that deliver the data in RDF • use of formal definitions for persons, places, concepts, events (URIs) - provide Linked and Open reference sources to link to (thesauri, taxonomies, etc) - build a Knowledge Graph for Dutch digital heritage - align and link related terms • add support for cross-domain discovery (EDM, Schema.org,...) => but how can this help our ‘Googling’ users to find our information?
- 11. Design issues... A tiny example...suppose a resource exists: <http://data.museum1.nl/object1> a nde:painting ; dcterms:subject <http://knowledgegraph.nde.nl/term1> . For ‘browsable Linked Data’ you would like to add the inverse relation: <http://knowledgegraph.nde.nl/term1> a skos:Concept ; skos:prefLabel "Flowers" ; skos:isSubjectOf <http://data.museum1.nl/object1> . TBL about ‘browsable linked data’ [1]: In practice, when data is stored in two documents, this means that any RDF statements which relate things in the two files must be repeated in each. [1]: https://www.w3.org/DesignIssues/LinkedData.html
- 12. Possible approaches to make it work • Just implement schema.org and let search engines do the work… • Aggregate all the related Linked Data resources: - collect datadumps and build a large triplestore - impressive work f.e. with LOD Laundromat • Use federated querying: - doesn’t work very well using regular triplestores - implement Linked Data Fragments, but what if we have 1500 nodes? - use logic for predicting most relevant endpoint [1]? [1]: Routing Memento Requests Using Binary Classifiers
- 13. Or should we build a registry? A decentralized infrastructure for the efficient management of resources in the web of data - Michalis Stefanidakis, Ioannis Papadakis - SWIM 2012
- 14. Register organisations, datasets and backlinks using DCAT, or DataID?
- 15. NDE’s strategy for resource discovery 1. improve the usability of the data at the source: - implement linked data principles - align references for people, places, events, concepts (URIs!) 2. build a knowledge graph for Dutch digital heritage 3. build a registry to link to the resources: - use formal definitions for organisations and datasets (DCAT/DataID) - send backlinks to registry (using Linked Data Notifications?) 4. support selective collection of data by federated querying (or harvesting): - use registry and knowledge graph for selecting the resources - use Linked Data Fragments for federated querying?
- 16. portal index data L O D website access to resources register align with URIs places, persons,... URIs datasets & object portal index query/ aggregate Knowledge Graph Registry collection registration systems (inter)national, regional or thematic portals collect resources (LOD or OAI) select relevant resources In a nutshell...
- 17. “Object ID” “term string” OAI OAI “Object ID” Term URI current situation Knowledge Graph enabled term publication via openSKOS LD LD Linked Data enabled Linked Data Platform compliant OAI based aggregator(s) Linked Data Platform LD LDLD API LD Object URI Term URI Object URI Term URI REGISTRY for organizations, datasets and links to objects KNOWLEDGE GRAPH for digital heritage collection registration systems (CRS) CRS CRS CRS CRS aggregators at national, regional, thematic level Term URIs Roadmap from OAI-PMH to LOD
- 18. TBL on Decentralization: This is a principle of the design of distributed systems, including societies. It points out that any single common point which is involved in any operation trends to limit the way the system scales, and produce a single point of complete failure. Interesting work: ● Interoperability and FAIRness [1] ● IPFS and Linked Open Data [2] ● https://discuss.ipfs.io/t/whos-using-ipfs-in-libraries-archives-and-museums/130 [1]: Interoperability and FAIRness through a novel combination of Web technologies - Mark D. Wilkinson, Ruben Verborgh et al. - April 2017 - https://peerj.com/articles/cs-110/ [2]: Sharing Linked Open Data over Peer-to-Peer Distributed File Systems: The Case of IPFS - Miguel-Angel Sicilia, Salvador Sánchez- Alonso - November 2016 - DOI: 10.1007/978-3-319-49157-8_1 A distributed network of digital heritage information??
- 20. Thank you for your attention! Please share your thoughts with me... [email protected] twitter/slideshare: ennomeijers https://github.com/netwerk-digitaal-erfgoed