ODDOL-Open Data Driven Online Learning
- Dataset publishers
- Conduct experiments and surveys
- Publish dataset with open licenses
- Researchers and Policymakers
- Use and analyse the datasets
- Test various data mining and machine learning algorithms, tools and visualization applications
- Publish interesting insights
- Online Learning Community
- Which dataset to use?
- Which algorithm to test?
- How to verify the published results?
- Which visualization is the best for this use case?
Contributions made during the weekend
- Architecture Design:
- Data Models:
- Information Required: Linking Datasets and Documents
- License of dataset
- Publisher of dataset
- Funding agency or sponsor behind the published datasets and results
- URL of the dataset and published document
- Data format used by the dataset
- Main subjects of the dataset as well as the published result
- Purpose behind use of dataset in the published document
- Works cited by the datasets (including other datasets)
- How to obtain these information?
- Curation by community members
- Automated annotation of published documents
- Where to store these information?
- Local instances (servers) including Wikibase instances
- Community managed data stores including Wikidata
- Mediawiki API
- Material design
- Shape Expressions (ShEx)
- Automated annotation linking datasets and results
- Periodic open dataset usage reports
- Analysis of results based on algorithms
- Ability to edit and describe a dataset and write a report directly on the platform
- Integration with other applications like Zenodo, Figshare, etc.
- Easy access on mobile devices