Linking Aid to the Sustainable Development Goals – a machine learning approach

Arnaud Pincet ; Shu Okabe ; Martin Pawelczyk

Organisation for Economic Co-operation and Development

关键词: Innovation;    Sectors;    Sustainable Development Goals;    Artificial Intelligence;    Text Mining;    Credit Reporting System;    Machine Learning;    Official Development Finance;   

DOI  :  https://doi.org/10.1787/4bdaeb8c-en

学科分类：社会科学、人文和艺术（综合）

来源: OECD iLibrary

【 摘 要 】

Official Development Assistance amounted USD 146.6 billions in 2017 but do we know how much of this aid contributed to the Sustainable Development Goals (SDGs)? And to what SDG in particular? This paper present a new methodology using machine learning designed to link project-based flows to the Sustainable Development Goals. It provide first estimates of DAC and non-DAC donors’ aid contribution for the goal and show that similar analysis can be done at the recipient level and for other type of textual database such as private sector reports; opening wide array for policy analysis.The methodology presented in this working paper uses semantic analysis of the text description of each project present in the Creditor Reporting System (CRS).