Analysis of the Malicious Bots Market

Kolomeets, Maxim; Chechulin, Andrey

doi:10.23919/fruct52173.2021.9435421

Cited by 12 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both of these data sets were compiled in the early days of the pandemic and, therefore, do not include any training data on recent COVID-19 misinformation and conspiracy theories. Additionally, we did not distinguish between bots or automated accounts and human users (Kolomeets and Chechulin, 2021). Future research might use more current training data and may distinguish between comments posted by humans and by bots to ameliorate any inaccuracies resulting from this.…”

Section: Limitationsmentioning

confidence: 88%

“…The emergence of new conspiracy theories, such as those about vaccines (Hornsey et al, 2021;Islam et al, 2021), might have increased the belief in misinformation about the pandemic and, accordingly, affective polarization. Second, new evidence shows that the YouTube search algorithm and other social media platforms amplify political extremism (Ribeiro et al, 2020;Huszár et al, 2021;Kolomeets and Chechulin, 2021). This kind of algorithmic bias may steer users toward higher levels of polarization as time passes.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

COVID-19 Induced Misinformation on YouTube: An Analysis of User Commentary

2022

View full text Add to dashboard Cite

Several scholars have demonstrated a positive link between political polarization and the resistance to COVID-19 prevention measures. At the same time, political polarization has also been associated with the spread of misinformation. This study investigates the theoretical linkages between polarization and misinformation and measures the flow of misinformation about COVID-19 in the comment sections of four popular YouTube channels for over 16 months using big data sources and methods. For the analysis, we downloaded about 3.5M English language YouTube comments posted in response to videos about the pandemic. We then classified the comments into one of the two following categories by applying a supervised Natural Language Processing classifier: (1)fake: comments that contain claims and speculation which are verifiably not true; and (2)legitimate:comments that do not fall into the fake category. The results show that the level of misinformation in YouTube comment sections has increased during the pandemic, that fake comments attract statistically more likes, and that the ratio of fake comments increased by 0.4% per month. These findings suggest that once introduced into an online discussion, misinformation potentially leads to an escalating spiral of misinformation comments, which undermines public policy. Overall, the results signal alarming pandemic-related misinformation and, potentially, rising levels of affective polarization. We place these results in context and point out the limitations of our approach.

show abstract

Section: Limitationsmentioning

confidence: 88%

Section: Discussionmentioning

confidence: 99%

COVID-19 Induced Misinformation on YouTube: An Analysis of User Commentary

2022

View full text Add to dashboard Cite

show abstract

“…To be more precise, it is essential to note various bot definitions, as researchers from different areas mention social bots differently [4]. Therefore, when we refer to bots in this study, we do not only mean automated accounts controlled by software but any account that someone can purchase to carry out controlled malicious activities, including hacked accounts, fake identities that are controlled by human operators, genuine-users who are paid to perform malicious actions, and any other types of bots that may emerge on the bot market [5]. This broad definition captures the full range of potential threats posed by malicious actors, regardless of the specific means by which they are carried out.…”

Section: Related Workmentioning

confidence: 99%

Experimental Evaluation: Can Humans Recognise Social Media Bots?

Kolomeets,

Tushkanova,

Desnitsky

et al. 2024

BDCC

Self Cite

View full text Add to dashboard Cite

This paper aims to test the hypothesis that the quality of social media bot detection systems based on supervised machine learning may not be as accurate as researchers claim, given that bots have become increasingly sophisticated, making it difficult for human annotators to detect them better than random selection. As a result, obtaining a ground-truth dataset with human annotation is not possible, which leads to supervised machine-learning models inheriting annotation errors. To test this hypothesis, we conducted an experiment where humans were tasked with recognizing malicious bots on the VKontakte social network. We then compared the “human” answers with the “ground-truth” bot labels (‘a bot’/‘not a bot’). Based on the experiment, we evaluated the bot detection efficiency of annotators in three scenarios typical for cybersecurity but differing in their detection difficulty as follows: (1) detection among random accounts, (2) detection among accounts of a social network ‘community’, and (3) detection among verified accounts. The study showed that humans could only detect simple bots in all three scenarios but could not detect more sophisticated ones (p-value = 0.05). The study also evaluates the limits of hypothetical and existing bot detection systems that leverage non-expert-labelled datasets as follows: the balanced accuracy of such systems can drop to 0.5 and lower, depending on bot complexity and detection scenario. The paper also describes the experiment design, collected datasets, statistical evaluation, and machine learning accuracy measures applied to support the results. In the discussion, we raise the question of using human labelling in bot detection systems and its potential cybersecurity issues. We also provide open access to the datasets used, experiment results, and software code for evaluating statistical and machine learning accuracy metrics used in this paper on GitHub.

show abstract

“…All major social media sites have a high prevalence of bots, and there is a substantial market for bots on TikTok [36]. We expect this will be an exceptionally prominent problem in the context of the Ukraine invasion, where many sides stand to gain or lose significantly from influencing public opinion.…”

Section: Botsmentioning

confidence: 99%

The Invasion of Ukraine Viewed through TikTok: A Dataset

Steel,

Parker,

Ruths

2023

Preprint

View full text Add to dashboard Cite

We present a dataset of video descriptions, comments, and user statistics, from the social media platform TikTok, centred around the invasion of Ukraine in 2022, an event that launched TikTok into the geopolitical arena. User activity on the platform around the invasion exposed myriad political behaviours and dynamics previously unexplored on this platform. To this end, we provide a mass-scale language and interaction dataset for further research into these processes. In this paper we conduct an initial investigation of language and social interaction dynamics, alongside an evaluation of bot detection on the platform. We have open-sourced the dataset and the library used to collect it to the public.

show abstract

Analysis of the Malicious Bots Market

Cited by 12 publications

References 5 publications

COVID-19 Induced Misinformation on YouTube: An Analysis of User Commentary

COVID-19 Induced Misinformation on YouTube: An Analysis of User Commentary

Experimental Evaluation: Can Humans Recognise Social Media Bots?

The Invasion of Ukraine Viewed through TikTok: A Dataset

Contact Info

Product

Resources

About