Hey all I have been doing a little bit of twitter bot hunting during these days of quarantine. My research has been focused mainly on the spanish far right, but honestly the same patterns can be observed all over the world.

Every social media user has possibly noticed how content is often driven by very strong emotions. Many posts range from extreme cute kittens to drama stories, but above all emotions it seems like online content is especially fueled by anger.

If we take a look at any controversial or political topic on twitter we get a glimpse of how the amplification machine of the far right operates.

At its core misinformation campaigns are built to spread a certain political agenda using specific language and amplifying only certain sources.

A few examples ..

I have been observing a number of accounts following this similar pattern.
There is a list of top talkers, usually these are politicians, journalists, online influencers. Whatever that means.

I have decided to check one of the top talkers for the spanish far right. For this users I checked a subset of the accounts retweeting them and got a subset of their timelines.

The user is alvisepf. You can check it out yourself.

I took only a number of tweets over a period of time. The archive is available here: https://www.hiro7.eu/static/files/twitter/query_result_2020-04-01T09_42_42.689784Z.csv

The idea was finding out what other accounts these users were tweeting.Most of the tweets are actually retweets to accounts on the spanish far right political spectrum.

These accounts are basically part of an amplification machine of the same group of top talkers for the spanish far right.

Hence we counted the number of retweets for account with the idea of extracting a distribution.

Here is the archive: https://www.hiro7.eu/static/files/twitter/query_result_2020-04-01T11_23_09.761662Z.csv

Here is a distribution per accounts:

We noticed that all these accounts always retweet the same news source. So here we compiled a list of tweets to these recurrent news sources from accounts that retweeted our top talker alvisepf.

Again archive: https://www.hiro7.eu/static/files/twitter/query_result_2020-04-05T17_07_24.229283Z.csv

Finally here is a complete archive of a set of tweets linking articles from these news source: https://www.hiro7.eu/static/files/twitter/users_tweeting_right_news_2020-04-05T16_51_11.893225Z.csv

What could be done to mitigate this? I think platforms are ultimately responsible.

I have heard people advocate for verified twitter accounts. I am not sure this is the way to go. A simple Proof of Work (PoW) mechanism
could well increase the cost of automating a large amount of highlights and retweets without compromising users' safety.

Finally, how can you run these analysis? There are a lot of open source libraries available. I particularly like this: https://github.com/twintproject 

I also have my own scripts using twitter apis.

And I use metabase ( https://metabase.com ) for analysis.

If you are a researcher. Please get in touch I'd be happy to open my DB and monitor other political scenarios around the world.

This thread is summarized on my blog: https://www.hiro7.eu/blog/2020-04-05-online-disinformation/