Experts: Elon Musk uses a wrong approach to count spam on Twitter
Dustin Moskovitz, the co-founder of Facebook, commented on the matter via his Twitter account. He pointed out that Musk’s method is not random and uses too few samples, leaving room for large errors.
He wrote: “Also, I feel like ‘doesn’t trust the Twitter staff to help pull the samples’ is its own kind of red flag.”
Christopher Bouzy, BotSentinel CEO, stated in an interview with CNBC, that his company’s analysis indicates that between 10% and 15% of Twitter accounts are likely to be “inauthentic,” which includes fakes, spammers and scammers, duplicates, and “single purpose hate accounts,” which target and harass people and others who spread disinformation.
BotSentinel is funded primarily through crowdfunding. It independently analyzes and identifies fake activity on Twitter using both machine learning software and human reviewers. Today, the company monitors over 2.5 million Twitter accounts. These are primarily English-language users.
Bouzy stated that Twitter does not have the ability to classify fake and spam accounts.
According to his warning, the number of fake accounts on Twitter can fluctuate depending on what topic is being discussed. BotSentinel found that more fake accounts are tweeting about cryptocurrency, politics, climate change, and other controversial topics than those who discuss non-controversial topics, such as kittens or origami.
CNBC interviewed Carl T. Bergstrom from the University of Washington. He co-authored a book that helps people understand data. CNBC also learned that sampling 100 followers of a single Twitter account are not sufficient to conduct “due diligence” in order to make a $44B acquisition.
Musk stated that 100 samples are far smaller than the usual sample size for social media researchers who study this type of thing. This approach would be fraught with selection bias, which Musk believes is the biggest problem.
Bergstrom sent a message to CNBC saying, “There is no reason to believe the followers of the official Twitter accounts are representative of all accounts on the platform.” Bots may be less likely to follow this account in order to avoid detection. They might be more likely to follow this account to make it seem legitimate. We don’t know. Musk may be doing something other than trolling us by using this ridiculous sampling scheme, but I don’t think so.
<< Previous