Social media networks like Twitter play a vital role in the dissemination of information. The Twittersphere is also known to house automated accounts. A study published by the Pew Research Center on April 9 revealed that 66 percent of tweeted links to popular websites came from bots.
Bots, as defined by CNET, are computer programs that talk like humans. Another example of a bot is Siri or Cortana. In other words, Siri and Cortana are intelligent personal assistants that perform automated tasks.
The Pew study and analysis set out to understand the role and nature of bots. In addition, researchers wanted to dig deeper as to why they are common on this social channel and how many links are being shared at any given time.
The think tank’s approach was to first create a list of 2,315 of the most visited and popular websites. Also, it examined more or less 1.2 million tweets. According to the Pew, it was specifically looking at tweets sent by English language users. This was done during a six-week period during the summer of 2017.
The findings from the organization included quite a few in great detail. One was the links contained within tweets to popular websites possess the characteristics of a bot. As stated by the Pew, where it gets more complex to comprehend this phenomenon of bots is when the share of a bot-created tweeted link is connected to a news website the automation is high.
“For example, an estimated 89 percent of tweeted links to popular aggregation sites that compile stories from around the web are posted by bots.”
Elsewhere, there is a small yet very active group of bots who are responsible for a significant share of links that feature prominent news and media websites.
The Twitter study goes on to compare the percentage distribution of links shared by the 500 most-active bots versus 500 most-active humans users. Naturally, bots outperformed their human counterparts. On the one hand, the analysis results showed bots were responsible for 22 percent of the tweeted links to popular sites whereas humans were responsible for a share of 6 percent. Particularly, this sharing of tweets by real people was significantly smaller.
During this study, the organization could not make identify whether these automated accounts had a political leaning. In essence, they were not able to identify or quantify accounts who had a conservative or liberal agenda.
As previously mentioned, the findings were based on a random sample of 1.2 million tweets from English Twitter users. The time frame of the study was from July 27 to Sept. 11, 2017.