The spread of misinformation and "fake news" online is a problem that has become increasingly concerning in recent years. Recently, researchers at Indiana University quantitatively "analyze[d] 14 million messages spreading 400 thousand articles on Twitter during ten months in 2016 and 2017" and found "evidence that social bots played a disproportionate role in spreading articles from low-credibility sources." In their article published in the open-access journal Nature Communications on November 20, 2018, the researchers explained how they used the open-source tools Hoaxy and Botometer (created and maintained by Indiana University) to calculate that while only six percent of the Twitter accounts in their random sample of the general population were identified as bots, those accounts were responsible for spreading over 30 percent of low-credibility content. The researchers also note that bot accounts may be targeting influential people with large numbers of followers in the hopes that these people will share the bots' content, lending it a facade of popularity and credibility. This detailed article includes several illustrative graphics as well as links to the code used to conduct its analysis and to supplementary material detailing the study's data sources and methodology.
Comments