People talk about anything on the web, especially on social media like Twitter.
On the microblogging site, cryptocurrency is one of the topics that seems to have no end to its debates. And here, the browser extension called 'Coinflict-Of-Interest' wants to help with this using algorithms to swift through tweets, followers and followees to understand users' particular cryptocurrency interest, if they have any.
Its creator Luke Childs explained the steps he usually use to determine how seriously he should take a cryptocurrency tweet.
This include understanding previous negative tweets, obvious bias from the past, and whether the person has criticized something towards which he or she is notionally biased.
"However this can be time consuming and it isn’t always easy to tell someone’s biases just by glancing at their profile," he argued.
As the ecosystem evolves, things can become nuanced, with people trying to boost or discredit a target by using innocent-looking tweets to deliver their intention.
And his 'Coinflict-Of-Interest' browser extension helps with that, allowing users to quickly get insights into user biases by providing a scaling blue bars underneath users' profile.
In the Twitter-verse where cryptocurrency space thrives. The entire cryptocurrency community is pretty much residing to either of the digital coins of their choice.
But in a social media where anything can happen, it's kind of a hellscape, where differences in investment portfolios often lead to in-fighting and disinformation. This can be a risk within the blockchain industry, as the line between cryptocurrency expert, venture capitalist and scammers is often blurred.
The browser extension here wants to help people avoid that, by exposing the political leanings of those who tweet about certain cryptocurrencies automatically, allowing users to gauge the legitimacy of their content.
"There are lots of people voicing their opinions on Crypto Twitter, however it’s often hard to work out if tweets are genuinely informative, or if the person has an ulterior motive," explained Childs, summarizing his intention in the extension's dedicated GitHub repository page.
Results for some figures are predicable, like Vitalik Buterin for example. The Ethereum co-founder is certainly more biased toward Ethereum. But some conclusions are more questionable.
"My algorithm does a reasonably good job of squashing out small chunks of followers from currencies you aren’t really aligned with,” said Childs. "But if you have a huge number of followers from another currency, it breaks down.”
Coinflict-of-Interest uses 'Hive.one' to determine bias.
Hive.one, according to a spokesperson, maps "how attention flows between Twitter accounts," to then organizes users into "clusters" of follower-followee relationships. It can ascertain the values of a given cluster by looking at the most dominant accounts associated with it.
Hive.one's algorithm is called 'Peoplerank', and it works similar to Google's PageRank. But instead of ranking websites and tracking links, it ranks identities and attention.
For Vitalik Buterin for example, the algorithm would assume him as a Ethereum fans. The extension will then translate his membership of certain clusters into bias. In this case, his bias in to Ethereum.
But the algorithm is flawed, and Hive.on admitted it.
And as for the extension, according to Childs, can only give feedback on users featured in its borrowed influencer lists.
Furthermore, if a user uses bots, has a small number of tweets, followers or followees, the tool can't really understand the context to determine the person's cryptocurrency interest.
But Childs thinks that the tool would improve were it to rely on “sentiment analysis.” That is, if it used AI to scan through influencers’ actual tweets to make judgements based on languages.
Coinflict-of-Interest is still a work in progress, added Childs.
Initially, the extension that is available for Opera, Google Chrome and Firefox, only accounts for Bitcoin, Ethereum, Ripple and Bitcoin Cash.