ATHENS -- Posts like a link to a viral video of dancing babies, followed by a breaking news alert, followed by a photo of someone's lunch make Twitter seem random.
But a University of Georgia researcher says the social network is completely predictable.
Itai Himelboim worked with the Pew Research Center and Social Media Foundation to track Twitter conversations. She broke down conversations into six predictable ways:
1. Polarized vs. Tight Crowd
"Polarized crowds are typically triggered by controversial political issues. Users interact with like-minded users, receive information from sources they agree with," Himelboim said. "It reminds me of high school politics- we don't talk with you, we don't listen to you, but we definitely talk about you."
The opposite is a tight crowd, where like-minded communities form around professional topics or hobby groups and exchange specific information.
2. Fragmented vs. Clustered
Fragmented (or brand clusters) are groups of people with no connection to one another tweeting about the same thing. Talk about celebrities falls into this category.
Groups of community clusters discuss global news. They form around a few hubs and can reveal the diversity of perspective on topics
3. Broadcast Network vs. Support Network
News media using Twitter to push out news creates a "broadcast network." News is retweeted by other people, setting agendas and starting conversations.
Support networks are the opposite, receiving communication instead of pushing it out. Businesses that use social media to manage customer issues interact with each individual problem.
The researchers behind the project say understanding how people connect can help determine public opinion and identify key influential people.