Researchers Uncover Method To Detect Online Trolls

in #science8 years ago

In a new study that was recently conducted by researchers at The University of Texas at San Antonio (UTSA), under leadership of Kim-Kwang Raymond Choo, it was uncovered that a certain algorithm can be used for detecting dishonesty online. Specifically, they utilized a statistical method in order to analyze the practice of astroturfing, and they sifted through a vast amount of writing samples in order to determine if they had originated from multiple accounts or whether they were a genuine posting from various individuals.

They say that it is difficult for authors to disguise their style of writing and so if those authors post from multiple accounts this sort of analysis would be able to pick up on that. It has the capability to find people who are using a number of different accounts in order to spread propaganda.

What Is Astroturfing ?

It is the practice of masking the original sponsor of a particular message, spreading it across what looks like multiple accounts, in order to make others believe that it had originated from grassroots participants.

The statistical analysis that the researchers employed in the study looks at context, punctuation, and choice of words, and it can successfully detect whether or not a group of posts are coming from just one person or whether they are the work of multiple authors. Choo and his team looked at writing samples from a number of infamous online commenters across a variety of sites.

They found that many of the comments which looked like they came from different sources, actually derived from only a few singular writers who are assumed to have multiple accounts.

The act of astroturfing works to create the illusion of consensus, when in reality that voice/opinion could likely be coming only from one source that is dishonestly disguised as many. Choo says that the practice is ethically questionable; though it has existed for as long as social media has been popular.

Unsurprisingly, many political figures have been accused of employing the practice in order to make it appear as though there is widespread support from the public for their platforms.

Choo says that they are now looking into whether or not the algorithm can be used to help in working to prevent plagiarism and contract cheating. They look forward to the prospect of this tool being used to help individuals make more informed choices and to assist them in trying to avoid online social manipulation and propaganda.

Pics:
southparkstudios
Sources:
https://en.wikipedia.org/wiki/Astroturfing
https://www.sciencedaily.com/releases/2016/10/161024095456.htm

Sort:  

Interesting, i hope we can find a effective method to fight these guyz.

Very interesting, anything that can help prevent plagiarism is a positive thing in my book.

Upvoted and reblogged ! All ready following! Loved your post, and I hope this really works soon to take care of these morons! Lol! ♨👌♨

Coin Marketplace

STEEM 0.15
TRX 0.16
JST 0.028
BTC 67340.80
ETH 2419.68
USDT 1.00
SBD 2.35