A research result that a small number of Twitter bots is enough to spread 'low-reliability content'


By Andre Pan

Diffusion of content of content that is delicious through SNS such as Facebook and Twitter has become a major problem in the Internet society. Research conducted on the influence of "fake news" on Twitter reveals that 31% of "less reliable content" is diffused by bot of only 6% of the total Twitter account It is on.

The spread of low-credibility content by social bots | Nature Communications
https://www.nature.com/articles/s41467-018-06930-7

'Relatively few' Twitter bots were needed to spread misinformation and overwhelm fact checkers, study finds
https://www.nbcnews.com/tech/tech-news/relatively-few-twitter-bots-were-needed-spread-misinformation-overwhelm-fact-n 939021

It is the team of researchers at the Indiana University in the United States that unveiled that a bot that occupies only a small percentage of the whole as a whole demonstrates great power in spreading fake news flowing on the platform. According to the published report, "Bot will amplify to the state that it is indistinguishable from fact-checked article by diffusing unreliable content," the diffusion effect of "fake news" by Bot is It is stated.

The study was conducted after Twitter shut down a lot of accounts in 2018 as "intervention from Russia affected the presidential election in 2016". The target was more than 14 million tweets posted from May 2016 to March 2017. From those tweets, it was linked to 400,000 articles, of which 380,569 articles were based on "low reliability source", and from the institution conducting the fact check "it was wrong It was said that it was classified as "it contains information." On the other hand, I also know that there were only 15,053 articles written in the source with articles that were approved by fact check.



In order to widely spread these articles, first of all it was adopted a method called "to increase diffusion power by using bots at the initial stage of diffusion". And once the article diffuses, it has been revealed that this time the bot was in the stage of "reaching a wider stratum with the influencer as the cornerstone".

In other words, first of all, by spreading the article to the maximum by the power of a number, after giving a certain presence to the article, using the diffusion power of influencer to disguise as if the article is reliable, In this research it has been shown that fake news was widely diffused in the world. Josh Russell, who is studying efforts to spread misinformation, said, "Twitter has made a great contribution to disinfecting bots reported by users, but we need to further improve the detection system," I am talking.

in Web Service, Posted by darkhorse_log