Twitter spam bots are a persistent and growing problem for the social platform where thousands of fake accounts can spread falsehoods to gullible human users.
But a study by the University of Southern California shows spam bots can be used to spread positivity, and even trigger good behaviour.
Co-authored by USC professor and computer scientist Emilio Ferrara, the study outlined a novel large-scale experiment that leveraged algorithm-driven Twitter accounts, what they dubbed social bots, to carry out coordinated tweets that shared positive hashtags.
This included behavioural prompts like #getyourflushot to encourage Twitter users to get vaccinated, #howmanypushups to encourage users to drop and give 20, and #highfiveastranger to engage in an act of kindness.
“We found that bots can be used to run interventions on social media that trigger or foster good behaviors,” said Ferrara in a press release. Ferrara and his team of researchers had previously studied online human-bot interactions.
Students from the Technical University of Denmark worked together to create a network of 39 social bots that were not only programmed to spread hashtags but could actually attract human followers. That came with its own set of challenges, making sure the botnets appeared to be human-like with a commendable follower to the following ratio and tweeting irregularly.
However, the bot network tweeted the hashtagged messages in a synchronized manner to actual Twitter users from a selected geographical area. They also retweeted human users who used the hashtag to further spread the positive messages, recording the behaviour of the target users, their interactions and how many times users were exposed to each message.
“The common approach is to have one broadcasting entity with many followers, but this study implies that it would be more effective to have multiple, decentralized bots share synchronized content,” Ferrara said. “Now we have seen empirically that when you are exposed to a given piece of information multiple times, your chances of adopting this information increase every time.”
The researchers explained in the study that coordinated interventions meant they could ensure the hashtags were in fact new to the Twitter world, deploy the bots to work together to expose users to the same hashtag message multiple times, and mitigate homophily–or the idea that birds of the same feather flock together on Twitter.
“While our work identifies the type of mechanism according to which information spreads from person to person, much work is still needed to discriminate which factors drive this phenomenon,” the study concluded.