APIs are an essential mechanism for delivering data to applications to support access, analysis, or machine learning. This is critical for data used to tackle global challenges, inform policy makers or the public, and support scientific progress. Such data includes official statistics, scientific data sourced across many disciplines, administrative data, and social networks.
Bringing it all together under a standard open APIs umbrella raises a wide variety of challenges that are not just technical in nature, and include metadata, standards, best practices, or governance. Being aware of all aspects of data management are essential for delivering high quality data APIs and content.
This presentation will highlight fundamental issues and challenges surrounding research and scientific open data APIs, describe ongoing efforts around related standard and best practices, and provide guidance for the establishment of modern data infrastructures.
The following topics will be covered:
- Machine Actionability: Why the current state of data limits usability and impairs machine intelligence, and how metadata and APIs play a critical role to improve upon this situation.
- Metadata: Which metadata standards should be used for open APIs? This will focus on ongoing global efforts to establish a Core Data Interoperability Framework and foster the establishment of a Global Open Science Cloud.
- FAIR: Findable-Accessible-Interoperable-Reusable (https://www.go-fair.org) data is a set of principles that have been widely endorsed by the research, scientific, and IT communities worldwide. How can such principles be reflected in API design?
- Implementation: How can API play a major role in the adoption and use of industry and domain data management standards? What are the practical integration challenges?
- Collaboration: Why establishing a strong dialog between information technologists and data scientists is essential to ensure robust open data API implementations.
Presentation slides
