Using Language Technologies to Automate the UNDP Rapid Integrated Assessment Mechanism in Serbian

Abstract

Rapid Integrated Assessment (RIA) is a United Nations Development Programme procedure involving a comparison between a country’s development policy documents and the UN-defined Sustainable Development Goals. In this paper, we present the Serbian AutoRIA system that automates this procedure in Serbian, a resource-limited yet morphologically rich language. We discuss the issues regarding the preprocessing of data for this task, and the general architecture and language-related specificities of the system. We also evaluate the performance effects of various system settings using the results of a previous, manually completed RIA procedure for Serbia.