Bob Yirka is a research scientist at Phys.org.
A team of researchers at the University of Washington's Center for an informed public have found that social media platform managers can reduce the spread of misinformation by combining just a few simple measures. The group used data from the 2020 presidential election to create a model that can predict misinformation. A summary of the findings by the group has been posted by the editors at Nature Human Behavior.
Over the past several years, people and organizations have recognized the danger of misinformation on social media platforms and have taken action. Managers of social network platforms have been put under pressure to reduce the spread of misinformation. They have taken measures such as banning bad actors and removing misleading posts. According to the researchers of the new study, such measures fail because they tend to show diminishing returns. Combining multiple measures is suggested by them.
The researchers obtained data from September 1 to December 15 of 2020 to find out how to combat misinformation. They pulled out 23 milliontweets about the U.S. presidential election. The software was used to find the ones that led to the events. A model similar to those used by epidemiologists was created by the researchers. They used the model to find scenarios that lead to the spread of information and misinformation. The best way to reduce the spread of misinformation is to apply all of the tools at the same time.
The researchers suggest that the spread of misinformation can be slowed by removing it as soon as it is identified, suspending repeat offenders, and putting warnings on posts that are not bad enough to remove, all at the same time. The researchers think that by combining these measures, misinformation could be reduced by over 50%.
More information: Joseph B. Bak-Coleman et al, Combining interventions to reduce the spread of viral misinformation, Nature Human Behaviour (2022). DOI: 10.1038/s41562-022-01388-6Nature Human Behavior says that modest interventions complement each other to reduce misinformation. There is a DOI of 10.1038/s41562.
Journal information: Nature Human BehaviourThere is a science network.