Andras Komaromi, Xiaomin Wu, Ran Pan, Yang Liu, Pablo Cisneros, Anchal Manocha, and Hiba El Oirghi
The International Monetary Fund (IMF) has expanded its online learning program, offering over 100 Massive Open Online Courses (MOOCs) to support economic and financial policymaking worldwide. This paper explores the application of Artificial Intelligence (AI), specifically Large Language Models (LLMs), to analyze qualitative feedback from participants in these courses. By fine-tuning a pre-trained LLM on expert-annotated text data, we develop models that efficiently classify open-ended survey responses with accuracy comparable to human coders. The models’ robust performance across multiple languages, including English, French, and Spanish, demonstrates its versatility. Key insights from the analysis include a preference for shorter, modular content, with variations across genders, and the significant impact of language barriers on learning outcomes. These and other findings from unstructured learner feedback inform the continuous improvement of the IMF's online courses, aligning with its capacity development goals to enhance economic and financial expertise globally.
This supplement provides background information on various aspects of capacity development (CD) for the main Board paper, The Fund’s Capacity Development Strategy—Better Policies through Stronger Institutions. It is divided into nine notes or sections, each focused on a different topic covered in the main paper. Section A explores the importance of institutions for growth, and the role the Fund can play in building institutions. Section B presents stylized facts about how the landscape for CD has changed since the late 1990s. Section C discusses the difficulties of analyzing CD data because of measurement issues. Section D provides a longer-term perspective on how Fund CD has responded to member needs. Section E contains information on previous efforts to prioritize CD, assesses Regional Strategy Notes (RSNs) and country pages, and suggests ways to strengthen RSNs, including by using the Fund’s surveillance products. Section F compares the technical assistance (TA) funding model proposed in the 2011