Original article in Kyiv Post: https://www.kyivpost.com/analysis/20577
Data was collected from the following sources:
Data collection code can be found in data.ipynb.
Note that in order to run the code you will need to populate the credentials directory with files fb_cookies.json, lj_cookies.json containing cookies for the corresponding sites, as well as put your Instagram username and password into the insta_credentials.json.
Be careful, Instagram and Facebook can temporarily ban you if you end up making too many requests.
I used BeautifulSoup to parse scrape posts from navalny.com and LiveJournal, snscrape for Telegram, facebook_scraper for Facebook and ensta for Instagram.
fbk_archive.csv contains aggregated data as a CSV table with the following columns:
author: Author of the posttext: Contents of the postdate: Date in the format%Y-%m-%durl: Link to the original postsource_type: Link to the source homepage
Author (Artur Chakhvadze) is not in any way affiliated with Navalny, FBK, ACF international, Kyiv Post or Oleksiy Borovskiy.
Anna Vellikok, a relative of Georgy Alburov, has promised to buy a beer to the Author as a token of gratitude for replicating the study.
Author does not hold any credentials whatsoever, contents of this repository should be judged on their own merits.