Lessons Learned from Building a Repository of Public Algorithms in Colombia
摘要
This chapter documents how the team of scholars built the new repository of public algorithms in Colombia and describes how the data was collected, processed, and organized. The chapter also explains the researchers’ main difficulties and solutions implemented. Finally, the chapter reflects on the challenges of fostering algorithmic transparency in a Global South country and offers recommendations for replicating this project in other countries. The database comprises 113 automated decision-making systems (ADMs) of 51 Colombian public bodies, characterizes each system with 40 variables, and was built with over 300 publicly available information sources. The ADMs are used by public organizations in Colombia to perform a wide range of functions and to support different types of state activities. Still, almost half of them are concentrated in the justice, education, and environment sectors.