Trolls, Bots and Everyone Else: Online Disinformation Campaigns and 2019 Presidential Elections in Ukraine

Urman, Aleksandra; Makhortykh, Mykola (25 October 2019). Trolls, Bots and Everyone Else: Online Disinformation Campaigns and 2019 Presidential Elections in Ukraine (Unpublished). In: 19th Annual Aleksanteri Conference: Technology, Culture and Society in the Eurasian Space. Helsinki, Finland. 23.10.-25.10.2019.

Full text not available from this repository. (Request a copy)

Today, online disinformation campaigns are increasingly employed to manipulate and alter public opinion in the context of elections. The use of coordinated disinformation efforts was traced in the recent elections in the US (Faris et al. 2017), France (Ferrara 2017) and Italy (Cresci et al. 2017). The purposes of these efforts varied from attacking specific candidates (Ferrara 2017) to forming negative attitudes towards certain social groups (Bennett & Livingston 2018). By doing so, disinformation campaigns corrode the foundations of democratic systems and increase societal polarization by dividing citizens along partisan lines (Tucker 2018). The research on online disinformation during elections is focused on two categories of agents: automated ones and human ones. The former are automated social media accounts (sock puppets) used to generate large volumes of content to support/attack candidates (Bessi & Ferrara 2016). The latter are human actors disseminating false information to condemn (i.e. “troll”) or praise (i.e. “elf”) candidates and their supporters (Bradshaw & Howard 2017). Until now, however, these two categories of agents are usually discussed separately, whereas in practice organized disinformation campaigns often involve both of them In our paper, we analyze the involvement of both automated and human agents in the online disinformation efforts during 2019 presidential elections in Ukraine. Two reasons motivate our choice of case study: firstly, as part of the ongoing Russian-Ukrainian conflict, Ukraine is frequently targeted by online disinformation campaigns sponsored by Russia (Mejias & Vokuev 2017). Considering the importance of presidential elections in Ukraine for the further course of the conflict it is highly probable that such campaigns would occur. Secondly, under the conditions of the ongoing information warfare, domestic Ukrainian actors increasingly adopt disinformation techniques to target their political opponents (Zhdanova & Orlova 2017) that further increases polarization in the Ukrainian society. To examine the interactions between human- and bot-produced disinformation and polarization in Ukraine, we are going to address the following research questions: How much online content was produced by bots and trolls compared with ordinary users in the case of specific candidates? How messages produced by bots and trolls differed in terms of the format and the purpose? And what was the impact of disinformation campaigns and if trolls or bots were more effective? To implement our research, we used Twitter REST API and captured tweets including the last names of the candidates with the highest electoral ratings in Cyrillic and Latin scripts between February 9 and April 30 2019. As an initial step for identifying disinformation agents, we used the anomalous frequency of posting and distinct positive/negative sentiment of content produced (Borra et al. 2017). We then differentiated between trolls and bots using logistic regression techniques based on user metadata and activity features (Ferrara 2017; Im et al. 2019). For comparing the content of messages produced by trolls and bots we used Latent Dirichlet Allocation model to identify the most common themes. Finally, for evaluating the impact of disinformation we employed Hawkes processes to assess the diffusion of manipulative content (Zannettou et al. 2018).

Item Type:

Conference or Workshop Item (Paper)

Division/Institute:

03 Faculty of Business, Economics and Social Sciences > Social Sciences > Institute of Mass Communication Studies

UniBE Contributor:

Urman, Aleksandra and Makhortykh, Mykola

Subjects:

000 Computer science, knowledge & systems > 070 News media, journalism & publishing
300 Social sciences, sociology & anthropology

Language:

English

Submitter:

Mykola Makhortykh

Date Deposited:

18 Dec 2019 16:40

Last Modified:

18 Dec 2019 16:40

Uncontrolled Keywords:

disinformation, trolls, elves, bots, Ukraine, elections, propaganda

URI:

https://boris.unibe.ch/id/eprint/135441

Actions (login required)

Edit item Edit item
Provide Feedback