January 22, 2021
DisQualified: Network Impact of Twitter’s Latest QAnon Enforcement
Graphika Team
Analysis of a QAnon network map produced by Graphika indicates that more than 60% of these Twitter accounts are no longer online. This network represents a set of 13,856 highly-connected accounts that were engaging with QAnon hashtags last spring; 8,859 of which are now inactive.
Read the full article hereGraphika Blog
Sort by:
Filter by Topic:
December 23, 2020
Graphika Global Insights: COVID-19 Vaccination Misinformation
In April 2020, Graphika published an inaugural report, The COVID-19 “Infodemic,” covering analysis of the global online conversation about the coronavirus pandemic from December 2019 through March 2020. At the time, we detailed how the COVID-19 pandemic had facilitated the convergence of health conspiracists, political conspiracists, and technology conspiracists on a global scale, while right-wing groups had played a significant role in creating and amplifying mis- and disinformation content.
Read the full article hereNovember 3, 2020
Graphika’s 24/7 ‘Peace Room’ Is Up And Running
Graphika is deeply committed to doing our part to protect the integrity of the 2020 United States presidential election. To this end, we have joined the Election Integrity Partnership to detect and mitigate the impact of attempts to prevent or deter people from voting or to delegitimize election results.
Read the full article hereSeptember 2, 2020
Q-Tips: Measuring the Mainstreaming of QAnon During the Pandemic
Journalists and researchers alike have observed the undeniable growth of conspiratorial content during the COVID-19 pandemic. Their important work shows the increased volume and frequency of sharing QAnon content across social media platforms over the last seven months. To help complete the picture, the Graphika team developed a new set of methods leveraging network analysis, topic modeling, and cultural bridging to identify content, users, topics and hashtags that connect otherwise disconnected communities.
Read the full article hereAugust 26, 2020
Dracula’s Botnet
A network of inauthentic Twitter accounts that exhibits strong signs of automated behavior has been amplifying the pro-China political spam network that Graphika has nicknamed “Spamouflage.” The accounts appear to be the work of a single entity, and all post apparently automated quotes from Bram Stoker’s “Dracula”; as such, we call this “Dracula’s Botnet.”
Read the full article hereAugust 24, 2020
Integrating Graph and Language Embedding Models—Graphika's Alex Ruch in Journal of Physics: Complexity
Can x2vec Save Lives? According to Graphika Labs team member Alex Ruch in his latest article published in the Journal of Physics: Complexity, the answer may be yes.
Read the full article hereJune 3, 2020
Briefing: China, Iran, and Russia State Accounts on U.S. Protests
State-controlled media outlets and official public diplomacy accounts in China, Iran, and Russia are focusing on the anti-racism protests in the United States, but they are primarily doing so in a way that furthers their existing narratives, rather than stoking American divisions.
Read the full article hereMarch 18, 2020
Pranksters or Foes? SMS-Based Crisis Disinformation
Previous viral SMS-based disinformation campaigns warn against mistaking online pranksters for foreign Russian trolls - and vice versa.
Read the full article hereMarch 16, 2020
Leveraging the Contagion Monitor™ Analytical Tool to Uncover Trends Before They Hit
Graphika’s Contagion Monitor™ analytical tool is a software solution that helps strategists collect and analyze social media data and generate sophisticated analytics to detect business relevant trends, gain more insights on emergent campaigns, and deliver exponential value for their brand.
Read the full article hereDecember 3, 2019
Graphika Report: UK Trade Leaks
A new report from Graphika finds that operators keen to hide their identities disseminated leaked UK/US trade documents in a similar fashion to Russian operation “Secondary Infektion”, exposed in June 2019.
Read the full article hereNovember 7, 2019
Deep Learning at Graphika: Scaling Network Maps with Heterogeneous Graph Embedding
Deep learning models for heterogeneous graphs allow us to overcome traditional analytical limits to efficiently scale our work at Graphika to tens of millions of nodes with hundreds of millions of edges.
Read the full article hereOctober 21, 2019
Graphika Report: The IRACopyPasta Campaign
On October 21, 2019, Facebook announced the takedown of 50 Instagram accounts posting about US social and political issues and the 2020 election. Facebook concluded that the operation “originated from Russia” and “showed some links to the Internet Research Agency (IRA)”, the Russian “troll farm” that previously targeted US audiences and the United States presidential election in 2016.
Read the full article hereSeptember 25, 2019
Graphika Report: Spamouflage
An active and prolific, but ultimately low-impact, cross-platform political spam network in Chinese boosted attacks on the Hong Kong protesters by using hijacked or fake accounts on YouTube, Twitter, and Facebook well into September 2019, an investigation by Graphika reveals.
Read the full article hereAugust 23, 2019
Graphika welcomes industry expert Ben Nimmo to the team
As the Graphika team continues to attract world-class talent across technology, marketing and insights, we are thrilled to welcome a new, noteworthy hire to our leadership team. This month, Ben Nimmo joins our team as our first-ever Director of Investigations.
Read the full article hereAugust 8, 2019
Hacking for the Greater Good: Empowering Technologists to Strengthen Digital Society
Camille François, Graphika’s Chief Innovation Officer visited Black Hat in Las Vegas to join a panel of experts discussing the intersection of technological advancement with the growing risks of cyber attacks and counter-applications.
Read the full article hereAugust 1, 2019
Graphika Labs Summer Retreat 2019: Innovating for the Future
On July 12th, 2019, Graphika Labs held its inaugural retreat in Boston, MA. We gathered to discuss our focus for the next twelve months and are excited to announce the five core pillars of Graphika’s research in 2019-2020.
Read the full article hereJune 25, 2019
Graphika Chief Innovation Officer Camille Francois Named as MIT Technology Review 2019 “Innovator Under 35”
Graphika's Chief Innovation Officer Camille Francois has been named to MIT Technology Review’s annual list of “Innovators Under 35” as a Visionary.
Read the full article hereJune 4, 2019
The Desus and Mero Story
One of the most elusive problems in social science is the prediction of hit phenomena — from transformative social movements to hit artists to virally marketed products. When Desus and Mero — a couple of comedians from the Bronx — got into a fight with DJ Envy, nobody knew that they were on the cusp of going viral, selling out a five-borough tour, and getting a show on Showtime — but Graphika was able to predict that one.
Read the full article hereIn the Media
How an Ex-YouTube Insider Investigated its Secret Algorithm
(The Guardian, February 2, 2018)
Graphika discovered over 513,000 Twitter accounts tweeted links to one or more YouTube-recommended videos that were boosted by vigorous social media campaigns leading up to the 2016 election, involving thousands of accounts controlled by political operatives, including a large number of bots.
Read the full report here
This Former Google Exec Talked to the Social Media Trolls the Russians Paid to Influence Elections — Here’s What She Learned
(CNBC, January 19, 2018)
Camille François, Graphika’s Chief Innovation Officer, shared first hand accounts of the stories of real troll farm workers and her own experience co-authoring a landmark report on Russian electoral influence operations for the Senate Select Intelligence Committee.
Read the full report here
The Insanity of NBA All-Star Voting
(ESPN, February 16, 2017)
Graphika sifted through more than 5 million tweets on behalf of ESPN, revealing a number of interesting findings about NBA All-Star voting, including 10 hyperactive bot accounts voting for Kawhi Leonard about 1,000 times per day.
Read the full report here