Bot detection using a single post on social media

Mohammad, Shad, Khan, Muhammad US, Ali, Mazhar, Liu, Leo, Shardlow, Matthew ORCID: https://orcid.org/0000-0003-1129-2750 and Nawaz, Raheel ORCID: https://orcid.org/0000-0001-9588-0052 (2019) Bot detection using a single post on social media. In: 2nd IEEE World Conference on Smart Trends in Systems, Security and Sustainability (IEEE WS4 2019), 30 July 2019 - 31 July 2019, London, UK.

Preview

Accepted Version Available under License In Copyright. Download (554kB) | Preview

Abstract

Recent studies of social media have made a unanimous conclusion that public opinions can be altered through systematic exploitation of social media using bot accounts. The existing bot detection methodologies utilize features of the accounts to label them as either bot or human. However, in this work, we propose a convolutional neural network (CNN) to identify the bot accounts using a single post on the social media. We have compared our results with an artificial neural network (ANN) trained on the features extracted from the accounts' profiles. Results have shown that bot accounts can be detected with 98.71% accuracy using CNN as compared to the 97.6% of ANN. Moreover, we have also proposed a model that combine both the techniques and have achieved 99.43% accuracy.