loader image
 

Brits or Bots? Take Us To Your Leaders…


Jeremy Corbyn:

We carried out some standard searches using Twitter Audit (a reliable first port of call which gives a general assessment of a Twitter account) and Truthnest.

Twitter Audit identified 9% of the @jeremycorbyn following as being suspicious, meaning there are signs of automation and inauthenticity, while Truthnest identified 6.8% of the following as inactive and 0.8% of the following as extremely active, both core indicators of inauthenticity.

You can view the full Truthnest report here:

We pulled dip sampling sets from both the inactive and extreme account groups and ran them through a tried and tested, stable process using Botometer (University of Indiana Social Media Observatory), Bot or Not (an AI tool by shinyapps), and Bot Sentinel to test and verify results. (Failure and accuracy rates for each were established during a 2018 study of Scottish Twitter commissioned by an MEP).

With the verification process of the dip sample complete, we can provide an estimate of fully automated (bot) and suspicious (human-managed) accounts which form part of the following of the @jeremycorbyn account.

The result sits within the normal ranges of over-identification by fully automated apps and sits within the lower bracket of the unhealthy range of suspicious activity around popular politician accounts across Europe, Australia, and North America.

We have not studied individual content posts or dip sampled non-following accounts for signs of suspicious activity.


Jo Swinson:

Following our analysis of Jeremy Corbyn’s Twitter account, we turned to Jo Swinson and carried out the standard searches using Twitter Audit and Truthnest.

Twitter Audit identified 6% of the @joswinson following as being suspicious, meaning there are signs of automation and inauthenticity, while Truthnest identified 7.9% of the following as inactive and 1% of the following as extremely active, both core indicators of inauthenticity.

You can view the full Truthnest report here:

We pulled dip sampling sets from both the inactive and extreme account groups and ran them through testing.

With the verification process of the dip sample complete, we can provide an estimate of fully automated (bot) and suspicious (human-managed) accounts which form part of the following of the @joswinson account.

The result sits within the normal ranges of over-identification by fully automated apps and sits within the lower bracket of the unhealthy range of suspicious activity around popular politician accounts across Europe, Australia, and North America.

We have not studied individual content posts or dip sampled non-following accounts for signs of suspicious activity.


Boris Johnson:

Following our analysis of Jeremy Corbyn and Jo Swinson’s Twitter accounts, we turned to Boris Johnson and carried out the same standard searches using Twitter Audit and Truthnest.

Twitter Audit identified 7% of the @borisjohnson following as being suspicious, meaning there are signs of automation and inauthenticity, while Truthnest identified 14.10% of the following as inactive and 1.40% of the following as extremely active, both core indicators of inauthenticity.

You can view the full Truthnest report here:

We pulled dip sampling sets from both the inactive and extreme account groups and ran them through testing.

With the verification process of the dip sample complete, we can provide an estimate of fully automated (bot) and suspicious (human-managed) accounts which form part of the following of the @borisjohnson account.

The result sits within the normal ranges of over-identification by fully automated apps and sits within the higher bracket of the unhealthy range of suspicious activity around popular politician accounts across Europe, Australia, and North America.

We have not studied individual content posts or dip sampled non-following accounts for signs of suspicious activity.

Ends.



Related Articles

Disinfonomics

This psychology not only continues to redefine our shared values, incorporating disinformation into our talking points and group behaviours, but profits the whole sector which both feeds upon and drives it. This is Disinfonomics.