Writing a bad memory about the phenomenon of misinformation that raged during the 2016 Russian presidential campaign, the scientists of the Pacific Northwest National Laboratory (PNNL) would like to pierce the subject by scrutinizing the behavior of the crypto-community on the site Reddit.
Understanding misinformation via cryptocurrency
Wishing to arm themselves against a possible scandal for the next Russian electoral campaign of 2020, the PNNL experts would like to follow a new track that promises to unveil the mechanisms of misinformation online.
Svitlana Volkova, PNNL Data Specialist, is confident that she calls crypto-currency “a very good proxy programfor misinformation”.
Certainly, it is difficult to unmask the true perpetrators of misinformation, who act, most of the time, in anonymity.
However, the analysis of the reactions and comments about crypto-currencies on the Reddit website shows a real similarity with facts observed in the past.
Svitlana Volkova and her teammates Emily Saldanha and Maria Glenski explored this track by conducting a study of tens of thousands of comments on Reddit forums about three crypto-currencies, namely Bitcoin (BTC), Ethereum (ETH) and Monero (XMR), between 2015 and 2018.
What the numbers say
This study showed that Bitcoin generates the greatest reactivity, with 3,600 daily comments posted against 500 for Ethereum and 380 for Monero.
However, discussions on Ethereum, which is both a cryptocurrency and a Blockchain development platform, tend to attract the curiosity of the crypto-community.
Monero is, according to researchers, more popular than Bitcoin, because of its high discretion in terms of anonymity of transactions.
As a result, the median value of the discussion time for Monero is the highest compared to that of Bitcoin and Ethereum.
By interpreting the data, the researchers were able to deduce that modeling would be under preparation:
“These social cues are very useful, and by incorporating them into the machine and deeper learning, we intend to build predictive models that address the causal relationships between different variables in order to explain the decision-making processes of the models “.
Do you think that the result of this study could help fight misinformation effectively? Comment in the comment section below.