Research question
The Sustainable Development Goals (SDGs) are the central guidelines for sustainable development. How can we determine the extent to which the goals are being applied? At the CEPS, we have developed a tool that can be used to analyze whether a text refers to the SDGs and if so, which ones. Based on the purpose or annual report of an organization, it is possible to analyze whether an organization is active in a certain SDG.

Role of AI
Keyword lists are a common method of measuring whether a text refers to SDGs. For example, if the words “poverty” and “alleviate” appear in a text, this text is assigned to SDG-1 (no poverty). There are several such expert-generated keyword lists. The problem with these lists is that they are subjective and lead to quite different results. Depending on the list used, different results are obtained. To counteract this and make the method more objective and accurate, we have trained a machine learning algorithm that optimally combines the predictions of the different lists.

Results
By applying machine learning, this model achieves higher accuracy and lower subjectivity (“bias”) than the individual keyword lists. We have published the model in an open-source R package (https://www.text2sdg.io) and it is already actively used for research.

Publications

• Wulff, D.U./ Meier, D.S./ Mata, R. (2024). Using novel data and ensemble models to improve automated labeling of Sustainable Development Goals. Sustain Sci: https://doi.org/10.1007/s11625-024-01516-3.
• Meier, D. S. (2023). The evolution of SDG-related third sector and public administration literature: an analysis and call for more SDG-related research. Sustainability: Science, Practice and Policy, 19 (1): https://doi.org/10.1080/15487733.2023.2236501.
• Meier, D. S./ Mata, R./ Wulff, D. U. (2021). Text2sdg: An open-source solution to monitoring sustainable development goals from text. arXiv preprint: https://arxiv.org/abs/2110.05856.