"Bot or Not" software checks tweets

That's what researchers at Carnegie Mellon University think they've found out. Tweets from bots are usually directed against corona restrictions. Telepolis reports that since January scientists have analyzed more than 200 million tweets in which corona virus or COVID-19 was discussed. They found that about 45% of the accounts behave like bots. They also catalogued 82 percent of the 50 most influential retweeters and 62 percent of the 1000 best retweeters as bots.

It is difficult to analyze whether a tweet is from a human or a bot. The machine learning based analysis tool "Botometer" from Indiana University Bloomington is known for controversial results. For this reason, the researchers use a self-developed tool called "Bot or Not", which evaluates about 1200 criteria.

Clear assignments of bots are when the tweets come from remote locations within a short time or fixed time sequences of the tweets. More difficult to evaluate are semanitic or grammatical evaluations of tweets.

But even if the number should be 50% too high, the number of tweets with "conspiracy theories" or clear fake news in the group of bots is frightening. A high number of reports about possible cures for COVID-19, about hospitals filled with mannequins, or tweets "explaining" the spread of the coronavirus through 5G radio masts, come from bots.

Read more https://www.heise.de/tp/features/Twitter-Bots-gegen-Corona-Massnahmen-4727379.html

Comment

If you like a crypted answer you may copy your public key into this field. (Optional) Public Key

To prevent the use of this form by spam robots, please enter the portrayed character set in the left picture below into the right field.

CC license   European Civil Liberties Network   Bundesfreiwilligendienst   We don't store your data   World Beyond War   Use Tor router   Use HTTPS   No Java   For Transparency