The following repository contains data to
- scrape the calendar information on sessions of the EP (
01_raw_data/calendar) - scrape the respective Table of Contents (
01_raw_data/tocs) - scrape the verbatim debate reports of the sessions (
01_raw_data/debates)
The debates are identified by:
- table of content points ending in (debate)
- excluding most table of content points that occured more than 10 times (adjounrment of the session, opening of the session, etc.)
- excluding some specific words related to proceedings that are not debates (votes, voting, minutes, etc.)
A rough visual inspection was done to confirm this. This resulted in about 22892 debates being collected.