Researchers at the University of Southern California (USC) have trained an artificial intelligence system to detect Twitter bots based on differences in the patterns of Tweeting activity of real and fake accounts.
The researchers analyzed two separate datasets of Twitter users, which were classified manually or by a pre-existing algorithm as either bot or human.
The manually verified data set included 8.4 million tweets from 3,500 human accounts, and 3.4 million tweets from 5,000 bots.
The team found that human users replied between four and five times more frequently to other tweets than bots did.
In addition, human users gradually become more interactive, with the fraction of replies increasing over the course of an hour-long Twitter session.
USC researcher Emilio Ferrara thinks the algorithm could complement bot-detection tools that analyze the language within posts.
From New Scientist
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA