In collaboration with PARIS21, OECD, and FAO's Office of the Chief Statistician (OCS), the Data Lab developed the MetroPolicy indicator that measures how decisions reflected in national planning documents (such as national development plans and poverty reduction strategy papers) are supported through the use of statistics, aimed at enhancing the synergy between good governance and official statistics.

MetroPolicy is made of innovative components that make it accurate and relevant, such as a machine learning text-mining algorithm that efficiently analyzes large documents to assess the extent to which such documents are evidence-informed. The algorithm is based on a set of keywords, a list of indicators from different international organizations, and a list of statistical terms related to the Sustainable Development Goals Agenda.

The strategy behind this tool is described in this presentation.