Working Papers

This paper proposes the use of synthetic training data generated by large language models
to improve machine learning SDG classifiers. It shows that supplementing existing training data with
synthetic data produced by the ChatGPT tool improves the performance of the SDGClassy classifier.
This addition of synthetic data is especially useful in building SDG classifiers given the limited availability
of properly labeled data and the complex, interconnected nature of the SDGs. Synthetic data thus enables
more effective machine-learning applications in this context.

In 2015, a set of Sustainable Development Goals (SDGs) will succeed the Millennium Development Goals as reference goals for international development for the period 2015-2030. Education was identified as a standalone goal (SDG4). Epistemic communities have documented a number of links between education and other SDG areas, and policy makers have long recognized many of them. Based on an exhaustive content analysis of 40 global reports, this paper examines how well such links are represented in flagship publications of the United Nations system. Taken together, the reports identify links between education and all the SDGs, with the notable exception of SDG 14 on oceans. For most of the…