Aggregation of cultural heritage datasets through the Web of Data
Download as PPTX, PDF0 likes106 views
The document discusses the challenges and requirements of metadata aggregation in the cultural heritage domain, emphasizing the need for improved interoperability and the use of standards such as schema.org and linked data. It outlines a pilot case study conducted within the Europeana network, which aims to automate metadata aggregation while adhering to the existing standards and functional requirements. The results suggest that while schema.org can effectively describe cultural heritage resources, further specifications may be needed for better integration and usability in metadata aggregation systems.
Aggregation of cultural heritage datasets through the Web of Data
- 1. SEMANTiCS 2018 - 14th International Conference on Semantic Systems September 2018
- 2. Title here CC BY-SA Outline ● Introduction and Motivation ● Our context for investigating the Web of Data ● Requirements for metadata aggregation ● Our pilot case study • Dataset descriptions for metadata aggregation ● Results and conclusions ● Ongoing and future work CC BY-SA Metadata aggregation of IIIF Resources at Europeana
- 3. Czech Republic, PD 1887, Uměleckoprůmyslové museum v Praze Preissig, Vojtech Coloured etchings Introduction Metadata aggregation in the cultural heritage domain
- 4. Title here CC BY-SA History ● Thousands of digital libraries exist, maintained by different organizations • hindering the discoverability, sharing and reuse of the cultural resources ● Federated search has been applied in some cases • ...but it did not scale, and had serious usability limitations ● Standard protocols for metadata aggregation gained more attention, mainly driven by preprint repositories • With OAI-PMH as the dominating standard ● Metadata aggregation centralizes the discoverability and access to resources to further promote the usage of the resources CC BY-SA
- 5. Title here CC BY-SA What kinds of technologies are we considering? ● The technological landscape and aggregation scale evolve… …and aggregation networks need to operate more efficiently to remain sustainable • What are the successors of OAI-PMH? ● Focus on adoption of technology that: • present low barriers for adoption by data providers • are used by Cultural Heritage institutions for other purposes •Search engine optimization •Linked data •Social web technologies CC BY-SA
- 6. Cristallisation ou Mouvement du temps, René Bord 1987, Bibliothèque Municipale De Lyon, public domain Our context for investigating the Web of Data
- 7. Title here CC BY-SA Motivation for studying the Web of Data ● Cultural heritage institutions are increasingly seeking wider interoperability on the Web ● We have identified linked data and Schema.org as options for innovating metadata aggregation. Use of these technologies may: • reuse expertise and deployed technology within the organizations • motivate the participation in aggregation networks CC BY-SA
- 8. Title here CC BY-SA Our earlier experiments ● Does crawling the Web (of documents) allow aggregators to reach structured metadata? ● Can cultural resources metadata be expressed in Schema.org? CC BY-SA
- 9. Title here CC BY-SA Reaching Structured data CC BY-SA ● We crawled 50.000 landing pages of resources from Europeana data providers ● We collected statistics about: • Usage of HTML5 meta tags, RDFa/RDFa lite, content negotiation/RDF • Data model in use: Dublin Core and Schema.org
- 10. Title here CC BY-SA Reaching structured data automatically is currently difficult ● In spite of the numerous activities, in cultural heritage, for making available linked data, reaching it through automated means is difficult • No wide support of HTTP content negotiation • ...and the structured metadata within the HTML pages is very limited or non-existent CC BY-SA
- 11. Title here CC BY-SA Representing cultural resources in Schema.org ● Can we find in Schema.org all the core modelling constructs of the EDM? • The resource that institutions provide metadata about schema:CreativeWork, schema:VisualArtwork, schema:Book, schema:Painting, schema:Sculpture, and schema:Product • The digital versions of the resource schema:MediaObject, schema:ImageObject, schema:VideoObject, schema:AudioObject • The contextual classes schema:Person, schema:Place and schema:Organization CC BY-SA
- 12. Title here CC BY-SA Representing cultural resources in Schema.org ● So all the core required modelling constructs are available, but... • … are cultural heritage institutions using them? • … can the very specific information needs of EDM/Europeana be achieved? • … do the original Schema.org patterns need to be further specified/restricted for metadata aggregation? CC BY-SA
- 13. Title here CC BY-SA CC BY-SA ● To evaluate, we used datasets where the same resources had metadata available in both Schema.org and EDM ● Schema.org (@data provider) ● EDM (@Digital Public Library of America) Representing cultural resources in Schema.org
- 14. Title here CC BY-SA Schema.org is suitable for describing cultural heritage resources ● The data providers prepared Schema.org metadata for Internet discovery, not specifically for cultural heritage aggregation ● ...despite this fact, the EDM metadata derived from Schema.org was close to fully suitable for aggregation by Europeana CC BY-SA
- 15. Title here CC BY-SA ...additional conclusions ● Specific semantics for aggregation and controlled vocabularies • Data providers may need to provide some specific properties, that are not being used for web search engines ● It may allow the description of additional types and characteristics not (yet) implemented in EDM • an opportunity for progressively improve the data available ● It will require recommendations and/or specifications regarding how data providers should provide their Schema.org metadata CC BY-SA
- 16. Title here CC BY-SA References of the studies CC BY-SA N. Freire, Martins B., Calado P. (2018) Availability of cultural heritage structured metadata in the World Wide Web. In proceedings of 22nd International Conference on Electronic Publishing ELPUB 2018. Available online: https://elpub.episciences.org/4608 N. Freire, Charles V., Isaac A. (2018) Evaluation of Schema.org for Aggregation of Cultural Heritage Metadata. In: Gangemi A. et al. (eds) The Semantic Web. ESWC 2018. Lecture Notes in Computer Science, vol 10843. Springer, Cham. doi:10.1007/978-3-319-93417-4_15. Preprint available online: https://2018.eswc-conferences.org/wp-content/uploads/2018/02/ESWC2018_paper_224.pdf
- 17. Cristallisation ou Mouvement du temps, René Bord 1987, Bibliothèque Municipale De Lyon, public domain Requirements for metadata aggregation
- 18. Title here CC BY-SA Constraints of our context ● The solution must fulfil the same functional requirements as the current aggregation solution of Europeana: • which is based on OAI-PMH and the Europeana Data Model ● The automation of metadata aggregation should be 100% based on Web of Data technology CC BY-SA
- 19. Title here CC BY-SA Requirements ● R1 - Data providers must be able to provide a LD resource of their dataset. ● R2 - All data transmissions between data providers and Europeana must be built on standard technologies of the Web of data ● R3 - Data providers must be able to provide machine readable licensing of their metadata • at the dataset level and also at the individual metadata record level ● R4 - Data providers must provide a machine-readable specification of how the dataset can be downloaded or harvested by Europeana • Two mechanisms may be used: RDF data dumps; or listings of the URI of the resources that are part of the dataset. ● R5 - EDM compliant metadata must be made available by the data provider. Alternatively, Schema.org metadata maybe used, as long as after conversion to EDM, it complies with the EDM schema requirements. CC BY-SA
- 20. Title here CC BY-SA Our Europeana pilot case study ● We addressed these requirements through a pilot case study within the Europeana network: • Including partner for the three roles of the Europeana network: •Data provider - The National Library of The Netherlands •Intermediary aggregator - The Dutch Digital Heritage Network •Central aggregator - Europeana Foundation ● We analyzed standards, created guidelines, developed software systems and defined data conversions ● Our paper presents the results on the description of datasets and their distributions ...applying vocabularies such as DCAT, VoID and Schema.org CC BY-SA
- 21. Title here CC BY-SA Sidenote: Linked Data activities from the National Library of the Netherlands CC BY-SA
- 22. Cristallisation ou Mouvement du temps, René Bord 1987, Bibliothèque Municipale De Lyon, public domain Dataset descriptions for metadata aggregation
- 23. Title here CC BY-SA Analyzed vocabularies for datasets CC BY-SA Vocabulary Req. R1 Provide an RDF resource Req. R3 Licensing Req. R4 Distribution of the dataset VoID - Vocabulary of Interlinked Datasets yes yes yes DCAT – Data Catalogue Vocabulary yes yes partial Schema.org yes yes partial EDM Datasets Profile yes no no ADMS - Asset Description Metadata Schema N/A - specific for semantic assets such as schemas, generic data models and vocabularies RDF Data Cube Vocabulary N/A - specific for multi-dimensional data Three requirements of metadata aggregation can be fulfilled by dataset descriptions. We identified and analyzed six candidates:
- 24. Title here CC BY-SA Three of the vocabularies were suitable CC BY-SA VoID - Vocabulary of Interlinked Datasets “VoID is an RDF Schema vocabulary for expressing metadata about RDF datasets. It is intended as a bridge between the publishers and users of RDF data, with applications ranging from data discovery to cataloging and archiving of datasets.” DCAT – Data Catalogue Vocabulary “DCAT is an RDF vocabulary designed to facilitate interoperability between data catalogs published on the Web. Publishers increase discoverability and enable applications easily to consume metadata from multiple catalogs. It further enables decentralized publishing of catalogs and facilitates federated dataset search across sites.” Schema.org The Schema.org vocabulary defines classes representing Datasets and their distribution
- 25. Title here CC BY-SA Supporting the usage of the three vocabularies Active communities in this area have provided alignments between the vocabularies: CC BY-SA Vocabularies Aligned Description DCAT and Schema.org From the W3C Dataset Exchange Working Group (DXWG). DCAT v1.1 (draft stage) includes the alignment of DCAT with Schema.org https://github.com/w3c/dxwg/blob/gh-pages/dcat/rdf/schema.ttl DCAT, ADMS and VoID The original work that lead to the creation of the Schema.org Dataset class. From a collaboration around the DCAT, ADMS and VoID vocabularies. http://www.w3.org/wiki/WebSchemas/Datasets DCAT and Schema.org From the W3C Spatial Data on the Web Working Group that included also the alignment with ISO 19115 - ‘Geographic information - Metadata’. https://webgate.ec.europa.eu/CITnet/stash/projects/ODCKAN/repos/dcat-ap-to-schema.org/browse
- 26. Title here CC BY-SA Example of a dataset description A case from the National Library of the Netherlands, using both Schema.org and VoID CC BY-SA
- 27. Title here CC BY-SA Example of a dataset description A made-up example of a DCAT description of a dataset with a downloadable distribution. CC BY-SA
- 28. Title here CC BY-SA ● We have put into practice a solution with low implementation barriers and use of existing expertise in CH institutions. ● Although the pilot is conducted just with three partners, we believe it provides a strong support for further research ● At present, the complete workflow of metadata aggregation based on the Web of Data was tested in the pilot CC BY-SA Conclusions on describing datasets
- 29. Ongoing and future work France, Public Domain Agence Rol. Agence photographique, Bibliothèque national de France Chat "regardant" à travers une longue-vue et autre chat perché dessus
- 30. Title here CC BY-SA Ongoing and future work ● The pilot is continuing. Currently it is addressing other requirements of the aggregation workflow ● It is also being extended to museum resources • taking its first steps (a cooperation with project CEMEC) CC BY-SA
- 31. Title here CC BY-SA Ongoing and future work Development of a prototype supporting all activities of the metadata aggregation workflow CC BY-SA
- 32. Thank you for your attention nuno.freire@tecnico.ulisboa.pt Netherlands, Public Domain 1660 - 1625, Rijksmuseum Anonymous Arrival of a Portuguese ship Acknowledgments Valentine Charles from Europeana Foundation Fundação para a Ciência e a Tecnologia (FCT): UID/CEC/50021/2013 European Commission contract number 30-CE-0885387/00-80.