Suchergebnisse

New developments in the analysis of FGV DAPP with data on the 2014 elections have shown the use of fake profiles to share content in the campaigns of Aécio Neves (PSDB), Dilma Rousseff (PT) , and Marina Silva (former PSB) . The analysis identified a botnet (bot network) with 699 profiles that shared content from Aécio Neves's and Marina Silva's campaigns. From the 699 accounts analyzed, 508 were created in 2 days (August 2 and 3, 2013) and published their last tweet between October 5 and November 2, 2014. During that period, the accounts generated 773.703 publications, an average of 419 publications per user. Some of the accounts are still active when drafting this report. The analysis of the content shared by these profiles showed that, among the most common links, are websites associated with Aécio's campaign and with the company Storm Security, provider of technology services. Based on the website of the Superior Electoral Court (TSE), it was identified that the Aécio Neves's campaign p aid R$168,000.00 in three installments to Storm Security. An investigation into "bots" that disseminated content of these domains also revealed a pattern of use of images linking them to the Eastern Europe, such as movie characters, places and captions in Russian. Figure 1, on page 9, illustrates a suspicious bot related to Aécio Neves in action. In the case of Dilma, there were searched 16,877 accounts for the links "mudamais.com.br" and "dilma.com.br" . In total, 430 bots shared the link for the Muda Mais website and 79 bots shared Dilma's website . A TSE decision that granted an injunction pointing the company Digital Polis – owner of the domain dilma.com.br – as the digital arm of Polis Propaganda : "the company that owns www.mudamais.com ", confirmed the relation between the domain and the campaign. Figure 2, on page 10, illustrates the performance of suspicious bots related to Dilma Rousseff.

English version available on: http://hdl.handle.net/10438/26229. ; In democratic elections, voting usually comprises two fundamental empowering aspects: electors vote for candidates because they agree with their projects and have a positive assessment of their background or party, and/or electors vote for candidates because of their charisma, honesty, and other values transmitted by candidates' own image (MUTZ, 2009). The dissemination of fake news affects both these motivations of a democratic vote. On one hand, the rational choice of projects and the assessment of candidates' background can be impaired by the overflow of possibly untrue information. On the other hand, untrue facts misrepresent the aforementioned features of candidates – either for or against their image. Besides, disseminating untrue information also reinforces ideological biases in each voter, because the evidence presented confirms ideas, values, and notions inherent to his/her points of view. 1 However, the fake news phenomenon cannot be fully understood based only on the false information element: the mode of dissemination is fundamental to determine the electoral impact. For that reason, social bots and real people share fake news in social media and communication applications such WhatsApp or even e-mail services. Social bots, when pretending to be real people sharing news and presenting opinions, have become a tool to generate a critic mass of sharing news in several online communication vehicles; consequently, they made some facts, either true or not, part of the mainstream narrative. Thus, what makes the current fake news phenomenon different from older forms of disinformation is precisely its massive scale and online circulation. ; Omidyar Network/Luminate