An exploration of the potential to use machine learning techniques to enhance the efficiency of analyzing, classifying, and synthesizing extensive amounts of text in evaluation research.
The resource offers an overview of machine learning and its pertinent applications within the evaluation field, investigating the potential to automate and therefore accelerate the extraction and classification of large text datasets. It is argued that this can be achieved through adequate training of the extraction tool and can provide evaluators with a robust analytical tool. The approach can be helpful in understanding the various factors influencing project success, identifying potential implementation challenges, and extracting valuable lessons for future endeavours.
Additionally, the paper showcases the Finance and Private Sector Evaluation Unit of the Independent Evaluation Group as a case study, highlighting the advantages of ML in text classification for evaluation purposes.
The paper concludes by presenting a summary of the experiment's outcomes and a brief exploration of potential next steps.
Bravo, L., Hagh, A., Joseph, R., Kambe, H., Xiang, X. & Vaessen, J. (2023). Machine Learning in Evaluative Synthesis: Lessons from Private Sector Evaluation in the World Bank Group. IEG Methods and Evaluation Capacity Development Working Paper Series. Independent Evaluation Group. Washington, DC: World Bank. https://ieg.worldbankgroup.org/sites/default/files/Data/Evaluation/files/methods_paper-machine_learning.pdf
These resources are part of the IEG methods papers series.
'Machine learning in evaluative synthesis: Lessons from private sector evaluation in the World Bank Group' is referenced in: