What is it about?
Social media, currently populated by hundreds of millions of users, represent efficient mechanisms for the communication and expression of the people who use them. However, they are also structures that present real incentives, including economic and political ones, to create algorithms that can impersonate real users who can manipulate the communication on these networks, in order to modify opinions, endorse (or destroy) reputations, manipulate markets or spread fake news. These algorithms are called social bots. It is essential to detect and restrain these agents, in order to allow communication in these media to occur organically, without the influence of third parties taking advantage of these means to spread their message, opinion or have any undue economic advantage . This paper investigated the main methods and challenges about social bots detection.
Featured Image
Photo by Jeremy Zero on Unsplash
Why is it important?
This work presents a systematic literature review (SLR) to find, interpret and evaluate studies related to bots detection mechanisms in social networks, in order to identify the main techniques developed, mechanisms used to validate these techniques, as well as their main advantages and disadvantages.
Perspectives
The detection of bots in social networks is far from being a solved problem, presenting challenges that demand the emergence of new approaches, since most studies fail to guarantee the effectiveness of the approach in the long term. Many techniques used are efficient for detection in datasets similar to the data that were used for the training of classifiers, not showing ways to adapt to the evolution of these malicious agents, requiring constant monitoring (and readaptation) in the classifiers used.
Dr. Luciano Antonio Digiampietri
Universidade de Sao Paulo Campus da Capital
Read the Original
This page is a summary of: Methods and Challenges in Social Bots Detection: A Systematic Review, June 2021, ACM (Association for Computing Machinery),
DOI: 10.1145/3466933.3466973.
You can read the full text:
Contributors
The following have contributed to this page
