Supporting materials for "Widespread use of National Academies consensus reports by the American public"

This repository contains supporting materials for "Widespread use of National Academies consensus reports by the American public." This includes (1) a table providing detailed information on each category into which comments left on NASEM reports were classified; (2) Python scripts for BERT implementation and associated *yml file; (3) BERT output files and model weights for 6-category and 64-category classification of user comments. This includes a compressed file containing the necessary: *.json, *.bin and vocab.txt files to use model weights for scientific replication. The specific file names are: config.json, special_tokens_map.json, tokenizer_config.json, vocab.txt, pytorch_model.bin