Have you used our tools such as @Botometer, Hoaxy, and Fakey? They will stop working soon once @Twitter discontinues free access to the API. This decision will affect many research projects and useful tools including those by @OSoMe_IU.
Our tools are free for all. Thousands of users daily leverage them to identify manipulation & study online narratives. If Twitter charges for the API, even at a very low price, we will have to terminate the tools or charge for use, excluding most users. #NoResearchWithoutAPI
2/
Without free @Twitter API, we won't be able to teach students in our classrooms how to study social media. Fakey, which helps students learn to recognize fake news, will also die, as will other tools (BotAmp, CoVaxxy, Midterm Elections dashboard, etc.)
W/out free API, research on social media manipulation will be impossible too. Examples: our past groundbreaking research on how (mis)information spreads in domains like the Occupy movement, Gezi Park protest, political elections. #NoResearchWithoutAPI
4/
Twitter data & OSoMe tools aided in finding the roots of Pizzagate & disinformation targeting White Helmets, taking down voter-suppression bots, and linking COVID-19 misinformation & vaccination hesitancy. reuters.com/article/us-usa…
Using @TwitterAPI data, we demonstrated the echo-chamber structure of information-diffusion networks on Twitter during the 2010 United States elections. Ten years later, this work received the @ICWSM Test of Time Award.
Using @TwitterAPI data we studied information virality patterns, in particular predicting what messages will go viral based on the structure of early diffusion networks and the role of competition for finite attention. nature.com/articles/srep0…
Using @Twitter API data we found the earliest evidence of astroturfing & social bots in 2010. We studied the cognitive, social & algorithmic vulnerabilities of social media platforms & users, and how bad actors exploit them.
Using @TwitterAPI we found that coordinated campaigns by inauthentic accounts threaten information integrity. We also demonstrated new forms of social media manipulation: bad actors grow influence networks & hide content with which they flood the network #NoResearchWithoutAPI
9/
NB: bots use the @Twitter API to WRITE content, but most research projects/tools use the API to READ data. To fight bad bots (
) @elonmusk could charge for writing, while keeping the read API free for researchers, journalists, etc! #NoResearchWithoutAPI
10/
We join other research centers in the Knight Research Network (@knightfdn) doing related work who posted similar threads articulating the many ways the @TwitterDev decision will harm research. @JosephineLukito's compilation:
In our white paper "Suspicious Twitter Activity around the Russian Invasion of #Ukraine" we present some preliminary evidence of suspicious activity obtained from analysis of 60 million tweets posted since February 1. Here is what we found:
The five most linked low-credibility sources include four Russian sources (rt, sputniknews, ria_ru, kremlin_ru) and zerohedge, a disinformation source amplifying Russian propaganda, according to U.S. intelligence sources.
Twitter caps our collection capabilities, as reflected in the flattening trend starting February 24. On the date of the full-scale invasion, we observed a significant increase in the creation of new accounts.
1/ Since 2010 we've been studying how to detect coordinated influence campaigns: one entity controls many accounts to amplify narratives and create false grassroots movements (aka astroturf).
2/ Here is an example of this type of manipulation, reported by @businessinsider: "How diehard Trump fans transformed their Twitter accounts into bots which spread conspiracies in a vast Russia-style disinformation network" businessinsider.com/power10-activi…
3/ And here is our recent paper in which we present a method to detect these kinds of coordinated manipulation networks: arxiv.org/abs/2001.05658