A GPT-based pipeline converted 235,000 scanned newspaper articles into a dataset of 78,685 violent events in Colombia, enabling descriptive analysis and a coca eradication regression that found no significant relationship.
World Drug Report 2023
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Using LLMs to create analytical datasets: A case study of reconstructing the historical memory of Colombia
A GPT-based pipeline converted 235,000 scanned newspaper articles into a dataset of 78,685 violent events in Colombia, enabling descriptive analysis and a coca eradication regression that found no significant relationship.