Disinformation and Social Bot Operations in the Run Up to the 2017 French Presidential Election (1707.00086v1)

Abstract: Recent accounts from researchers, journalists, as well as federal investigators, reached a unanimous conclusion: social media are systematically exploited to manipulate and alter public opinion. Some disinformation campaigns have been coordinated by means of bots, social media accounts controlled by computer scripts that try to disguise themselves as legitimate human users. In this study, we describe one such operation occurred in the run up to the 2017 French presidential election. We collected a massive Twitter dataset of nearly 17 million posts occurred between April 27 and May 7, 2017 (Election Day). We then set to study the MacronLeaks disinformation campaign: By leveraging a mix of machine learning and cognitive behavioral modeling techniques, we separated humans from bots, and then studied the activities of the two groups taken independently, as well as their interplay. We provide a characterization of both the bots and the users who engaged with them and oppose it to those users who didn't. Prior interests of disinformation adopters pinpoint to the reasons of the scarce success of this campaign: the users who engaged with MacronLeaks are mostly foreigners with a preexisting interest in alt-right topics and alternative news media, rather than French users with diverse political views. Concluding, anomalous account usage patterns suggest the possible existence of a black-market for reusable political disinformation bots.

• The paper identifies that 18% of accounts in the MacronLeaks campaign were bots, emphasizing their role in spreading disinformation.
• The study uses a data-driven approach combining logistic regression and cognitive behavioral modeling to analyze 17 million tweets over 11 days.
• The findings reveal that bot-generated tweets preceded human posts and targeted an international alt-right audience, misaligned with French voter engagement.

Analysis of Disinformation and Social Bot Operations: The Case of the 2017 French Presidential Election

The paper under scrutiny provides an in-depth analysis of the disinformation campaign known as "MacronLeaks" that occurred during the 2017 French presidential election. The paper capitalizes on a vast dataset consisting of around 17 million tweets collected over a period of 11 days leading up to the May 7, 2017, election day. The primary objective of the research is to illuminate the dynamics of social bots and the spread of disinformation in the context of a pivotal political event.

Methodological Approach

The research employs a data-driven approach incorporating machine learning techniques and cognitive behavioral modeling for bot detection and classification. This approach involves the extraction and examination of specific Twitter data related to election discourse, focusing particularly on tweets containing hashtags associated with the MacronLeaks campaign. A logistic regression model was utilized to classify accounts as bots or humans, based on a range of metadata features including tweet frequency and follower-followee ratios.

Key Findings

1. Bot Participation and Influence:
   • The paper identified 18% of accounts engaged in the MacronLeaks campaign as bots, a finding consistent with previous studies on political disinformation. These bots were responsible for amplifying disinformation narratives, notably in the days preceding the election.
   • The paper hypothesizes the existence of a black-market for reusable disinformation bots, evident from the reactivation of accounts used in the 2016 U.S. presidential election.
2. Audience Composition:
   • A significant portion of the users engaging with MacronLeaks were from the alt-right community and resided outside France, predominantly in English-speaking regions. This demographic discrepancy suggests the limited impact of the campaign on French voters.
3. Disinformation Dynamics:
   • The temporal analysis illustrated that bot-generated tweets often preceded human-generated posts, suggesting a catalytic role for bots in disinformation cascades.
   • The paper highlights a distinct characterization of disinformation narratives tailored for an international, English-speaking audience, which contrasts with the dominantly French language and nuanced political discourse among general election discussions.
4. Content Analysis:
   • Analysis of URL sharing patterns identified a bias towards hyper-partisan and alt-right news sources during the MacronLeaks campaign. This provides insight into the media ecosystem leveraged to propagate disinformation.

Implications and Future Directions

The implications of this paper extend to the broader understanding of how computational propaganda operates in political contexts. The paper underscores the role of social bots in polluting political discourse and manipulating public opinion, though in this case, with limited success due to the misalignment between the campaign's target audience and the voting populace.

The potential existence of a market for reusable political bots raises concerns about future elections and events, emphasizing a need for enhanced detection methods and preventive strategies. Subsequent research could focus on comparative analyses of disinformation campaigns across different geopolitical contexts, aimed at developing robust frameworks for identifying and mitigating such influence operations.

In conclusion, this paper contributes significant empirical evidence to the growing body of literature on social media manipulation and offers methodologies that can be employed in detecting and analyzing bot activities in future studies. The findings provide a foundational understanding of the mechanics behind disinformation campaigns, with both theoretical insights and practical security implications for safeguarding democratic processes.