Hierarchy of skill categories AI.
OECD.AI visualisations using data from these data sources were developed by the AI Lab at the Slovenian Jožef Stefan Institute (JSI). Data sources and methodologies on OECD.AI may differ from those previously used for OECD research, and therefore results may vary. Because methods for measuring AI are still evolving, there remain definitional and other issues that can influence empirical results. Efforts to develop clear definitions, taxonomies and classifications are underway, as are efforts to compile accurate and comparable indicators.
Given the current evolving situation, showing trends in a timely manner based on transparent methodologies by partner institutions can be of value to policy makers.
The “Trends & data” section of OECD.AI aims to help policy makers develop, implement and improve policies for AI. To do so, OECD.AI showcases AI-related data and metrics over time from as many high quality sources and vantage points as possible to allow countries and stakeholders to compare policy responses, engage in international co-operation, monitor progress and develop best practices.
The below notes aim to provide a full and transparent account of the sources and methodology used to construct the “Live data” visualisations on OECD.AI:
- AI news: Event Registry
- AI research: Microsoft Academic Graph | Elsevier (Scopus)
- AI jobs and skills: LinkedIn | Adzuna
- Investments in AI: Preqin
- AI software development: GitHub
- AI education: Studyportals | Coursera
- AI search trends: Google Trends
- AI demographics: Stack Overflow Survey
- AI knowledge transfer: Stack Overflow