Datastructuring Fragmented Lives: A Probabilistic Record Linkage to Historical Data

My doctoral research explores livestock expansion in the 18th-century Pantanal, a contested frontier between the Spanish and Portuguese empires. I argue this peripheral region was integrated into a vast economic zone under an indigenous geopolitical order. Utilizing a multidimensional database of 115,545 records from colonial Southern Brazil, my study currently faces the challenge of pervasive homonyms. Through the CDH fellowship, I will employ the Python package Splink to develop a probabilistic record linkage model. This technical intervention will generate unique identifiers (UIDs) for individuals, migrating fragmented biographical data into a unified SQLite environment. This will later allow me to map historical locomotion, production, and land usage, as well as transform my dataset  into a publishable, open-source resource.