Published in

MDPI, Applied Sciences, 9(11), p. 4105, 2021

DOI: 10.3390/app11094105

Links

Tools

Export citation

Search in Google Scholar

Bot Datasets on Twitter: Analysis and Challenges

This paper is made freely available by the publisher.
This paper is made freely available by the publisher.

Full text: Download

Green circle
Preprint: archiving allowed
Green circle
Postprint: archiving allowed
Green circle
Published version: archiving allowed
Data provided by SHERPA/RoMEO

Abstract

The reach and influence of social networks over modern society and its functioning have created new challenges and opportunities to prevent the misuse or tampering of such powerful tools of social interaction. Twitter, a social networking service that specializes in online news and information exchange involving billions of users world-wide, has been infested by bots for several years. In this paper, we analyze both public and private databases from the literature of bot detection on Twitter. We summarize their advantages, disadvantages, and differences, recommending which is more suitable to work with depending on the necessities of the researcher. From this analysis, we present five distinct behaviors in automated accounts exhibited across all the bot datasets analyzed from these databases. We measure their level of presence in each dataset using a radar chart for visual comparison. Finally, we identify four challenges that researchers of bot detection on Twitter have to face when using these databases from the literature.