What do you make of mostly dormant Twitter accounts that suddenly become active and pro-trump

[xor]

Whenever I see a very very pro-trump account that fits a certain template, I like to use the advanced twitter search to pull up their early tweets. Some accounts that have been around for years often times have nothing up until they began their Pro-Trump tweets. These, I assume, were likely cleared and then had the name changed. Then there are older accounts that were mostly inactive, but often times had some non-political tweets on them, but then within the last year or two they became very very active posting pro-trump and conspiracy junk constantly.

Here's an example of one over a two year period right before constant posting of political stuff started in Nov 2016.
https://twitter.com/search?l=&q=from%3AGenevagirl81%20since%3A2014-11-01%20until%3A2016-11-01

I see this sort of stuff all the time. Are these crazy people who just became politically active on twitter because Trump is tweeting obsessed idiot? I'm sure some are, but a good portion of the most popular and loud pro-trump "voices" on twitter follow this trend. So, there is the possibility they are bots or at least controlled propaganda accounts. But how did they come about? Were they accounts that were taken over? Were they pre-staged accounts with the ultimate goal to go active? Of course some of these accounts are real people,a and they are not all bots. Being called a bot is now similar to being called a troll. People misuse the term and call anyone who disagrees with them a bot (or troll). I've been called a bot many times for reasons I don't fully understand.

One time I did a search on a pro-trump account that was pushing fake news stories. About five years prior they were posting what I would consider to be far leftist and I would say very anti-US posts. I called them out on it and they quickly deleted their entire account. That's part of why I am always really skeptical of anyone who waves the liberal flag, but calls for violence and/or uses very divisive language.

[Basement Beat]

1. That's interesting... would be great if there was

more data on this. Wonder if there's a tech group that has some numbers on this. Just to see how wide of the scope is.

[xor]

2. Yeah, I would be really interested in seeing what data Twitter has.

They are probably the only ones who can really dig in deep with the data to figure stuff out. Which I'm assuming they do given that we occasionally hear of them purging bot accounts. I know there are various services out there that will take a twitter handler and give it a bot score, but I'm skeptical of those since they tend to be fairly opaque in their methods. So even though they may be using fancy machine learning stuff, I would think those still need to be trained using the fairly limited publicly accessible data.

As creepy as that stuff can be, I would love to be able to work with that data. I'm by no means any sort of expert in that arena, but it would be neat to be able to work on it.