Design of a Data-driven Intervention Dashboard for SDG Localization

Design of a Data-driven Intervention Dashboard for SDG Localization

Pooja Bassin, Abraham G K, Srinath Srinivasa

Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence
Demo Track. Pages 8606-8609. https://doi.org/10.24963/ijcai.2024/990

The localization problem of the United Nations Sustainable Development Goals (SDGs) involves adopting strategies that are in tune with local conditions, to achieve a given SDG target. However, even within a given region, localized conditions may vary drastically. With increasing amounts of Open Government Data (OGD) being available, there is an opportunity to systematically address the localization problem by using predictive and prescriptive modeling techniques. This work presents a predictive and prescriptive modeling dashboard for the SDG indicator maternal deaths (MD) for the Indian state of Karnataka. The dashboard was created by examining a vast set of data points to focus on four factors that showed high correlations with that of MD. We then construct a multivariate linear regression model to showcase the differential impact that a given factor has on the indicator and identify prescribed values for different factors to achieve a given target value of the indicator. Finally, a budget allocation dashboard is also provided that helps policymakers allocate budgets to specific schemes to help operationalize these changes. This dashboard was built by combining data coming from five different OGD sources.
Keywords:
Multidisciplinary Topics and Applications: MDA: Energy, environment and sustainability
Data Mining: DM: Data visualization
Machine Learning: ML: Applications
Machine Learning: ML: Regression
Multidisciplinary Topics and Applications: MDA: Social sciences