Model-driven engineering and predictive analytics for implementation of sustainable development goals

Abstract

Improving the quality of implementation of sustainable development plans remains a concern among involved actors. Measurement, monitoring and evaluation are the three cardinal phases that any sustainable development plan is subjected to in order to guarantee successful implementation. The less-than-impressive implementation of previous global sustainable plans such as the Millennium Development Goals (MDGs) mean that there is need to explore a better approach to measuring, monitoring, and evaluating of plans at national and sub-national levels. The failure of the MDGs (2000-2015) has been partly blamed on inadequate use of data expressed in terms availability on one hand, and inability to detect and predict patterns in the available data on the other hand. In this present work, we improve on measurement and monitoring by promoting understanding and collaboration among stakeholders using model-driven engineering (MDE). MDE simplifies and demystifies system development concepts. This is achieved by graphically demonstrating entities and relationships and quantifying them. We also propose the integration of predictive analytics in the evaluation of sustainable development objectives using deep learning neural networks (DLNN). This machine learning technique detects patterns in existing data and predicts outcome of future data using patterns learnt. The combination of these contributions in improving measurement, monitoring and evaluation is aimed at improving the implementation of present and future global plans such as the Sustainable Development Goals (2015 -2030).