Data science for sustainability

Data science for sustainability

Graph partitioning and application to the study of the relationships between the 17 Sustainable Development Goals

In data science, graph partitioning allows structuring and summarizing the information contained in a network by automatically discovering disjoint groups within which the nodes are highly interconnected. Graph partitioning is a combinatorial problem that has been studied by many papers and which has been applied in numerous fields. We describe our contribution on this topic and we apply our new method to the study of the relationships between the 17 Sustainable Development Goals. The results allow us to underline different kinds of axis which can be either “composite” or “unitary” when analyzing the sustainable development of countries.

Read the blog article in French

Reference: Ah-Pine, J. (2022). Learning doubly stochastic and nearly idempotent affinity matrix for graph-based clustering. European Journal of Operational Research, 299(3), 1069‑1078. https://doi.org/10.1016/j.ejor.2021.12.034.