3. Semantic publishing for OpenAIRE
• Linked entities
• Beyond a flat data model – CERIF compliant
• Overlapping efforts in data modelling basic entities
• Using multiple identifier schemes
• Discipline specific best practices (DOIs, PIDs, URI/URN’s, db ids, …)
• Contextualizing by relationships
• Multiple types and vocabularies
Publications in context
The future of data publishing. Oxford May 22, 2013 3
4. Semantic enrichment services
• Citation discovery
• Text mining – lots of it…
• Discipline specific algorithms
• Classification
• Supervised
• Discipline specific vocabularies – library oriented
• Training sets – hard to find
• Unsupervised classification
• Interdisciplinary complexity
• Finding trends
Citation, classification, clustering
The future of data publishing. Oxford May 22, 2013 4
5. Zenodo
• Metadata general enough not to capture discipline semantics
• Different types of material
• Supplementary data or …?
• Context in relation to funding and publication
• Community regulated quality
• To be linked to OpenAIRE text mining services for metadata
enrichment
An all purpose data repository – www.zenodo.org
The future of data publishing. Oxford May 22, 2013 5
6. Challenges
• Implementation of guidelines/standards
• OpenAIRE guidelines for literature, data, CRIS
• Global alignment and adoption (RDA, WDS, W3C, …)
• Uniform vocabularies to support
• Interdisciplinary classification
• Multilinguality (e.g., EUROVOC)
• Links to other domains
• Links to other domains
• Mapping of data models (DCAT, LOM, …)
• Existing projects (e.g., fp7 ENGAGE)
• Tools for semantic enrichment at publishing time
The future of data publishing. Oxford May 22, 2013 6