mercredi 23 août 2017

Data mining finds a TAXONOMY OF TROLLS

Anything discussing data mining, characteristics of the alt-right on social media, social media trolls and bots are acceptable in this discussion. Just try to stick to the data mined characteristics rather than the politics of the alt-right which we already have plenty of threads on. Thanks.

This post in the Trump thread really needed its own discussion:
Quote:

Originally Posted by halleyscomet (Post 11969322)

This is the original article: Linguistic data analysis of 3 billion Reddit comments shows the alt-right is getting stronger
Quote:

The alt-right isn’t one group. They don’t have one coherent identity. Rather, they’re a loose collection of people from disparate backgrounds who would never normally interact: bored teenagers, gamers, men’s rights activists, conspiracy theorists and, yes, white nationalists and neo-Nazis. But thanks to the internet, they’re beginning to form a cohesive group identity. And I have the data to prove it.
Welcome to the Brave New World of data mining.

The other interesting thing in that analysis besides establishing a taxonomy of trolls is the expanding use of data mining.

I also wonder which of those groups or percentage of each might be bots, especially given the data mining looked at frequently used words and phrases.


Here are some additional articles on bots involved in the social media surrounding the alt-right.
How to Spot a Social Bot on Twitter
Quote:

Back in 2011, a team from Texas A&M University carried out a cyber sting to trap nonhuman Twitter users that were polluting the Twittersphere with spam. Their approach was to set up “honeypot” accounts which posted nonsensical content that no human user would ever be interested in. Any account that retweeted this content, or friended the owner, must surely be a nonhuman user known as a social bot.

The team set up 60 honeypots and harvested some 36,000 potential social bot accounts. The result surprised many observers because of the sheer number of nonhuman accounts that were active. These bots were generally unsophisticated and simply retweeted more or less any content they came across.

Since then, social bots have become significantly more advanced. They search social networks for popular and influential people, follow them and capture their attention by sending them messages. These bots can identify keywords and find content accordingly and some can even answer inquiries using natural language algorithms.

That makes identifying social bots much more difficult. But today, Emilio Ferrara and pals at Indiana University in Bloomington, say they have developed a way to spot sophisticated social bots and distinguish them from ordinary human users.
Research links pro-Trump, anti-Macron Twitter bots
Quote:

Through an analysis of nearly 17 million Twitter posts made between April 27 and May 7, Ferrara and his team turned up roughly 100,000 Twitter users engaged in conversation about the leaked emails from Macron’s campaign, about 18,000 of which they determined to be bots — social media accounts controlled by computer scripts that masquerade as human users.

A large majority of the users engaged in the conversation, he said, had a history of showing support for Trump, Republican or far-right narratives. Many of them were also English speakers, he said.

The scale of the bot operation in the French election, Ferrara observed, was much smaller than that related to the 2016 presidential election. He detected between 400,000 and half a million Twitter bots engaged in the U.S. political discussion, most of them pushing pro-Trump or far-right narratives. A minority of the suspected bots supported Democrat Hillary Clinton.
Sometimes crowdsourcing can efficiently mine data:
Donald Trump’s Twitter Following Might Include More Than 4 Million Bots
Quote:

Ever since assuming office, Trump had added almost 7 million new followers, but over 4 million of those new accounts didn’t feature a profile picture, but rather an egg avatar, which is a usual sign of a fake account. According to Twitter Audit, only 55 percent of Trump’s Twitter followers are real accounts, meaning 15 million of his #MAGA followers are phony.
You can bet if lack of an image identifies a bot, that bot creators will quickly remedy that tell.... but I digress.


via International Skeptics Forum http://ift.tt/2vpGUac

Aucun commentaire:

Enregistrer un commentaire