According to the OpenFDA problem definition, it says:
Level 1: Create visualization (word cloud) of boxed warnings
Level 2: Develop a language model to categorize the language in the warning sections of SPLs
Are the data elements to be used for these two different? That is, should I pick up "boxed_warning" for Level 1, and "warnings" for Level 2 from the API output?
Thanks,
J