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.

While migration and population displacement has always been part of the human experience, the context within which it occurs today has materially changed. Migration has become an important part of economic globalization and closely related to countries´ development process. Conflicts, poverty, natural disasters and climate events are also forcing people to migrate in an ever-increasing number. For many low-income countries with large number of internally-displaced people, on the other hand, the high economic costs are making it more difficult for them to invest in SDG implementation. Developing countries also host most of the externally-displaced people at high economic costs, which…