Twitter communities are definable by the language they use
Social media tools are a great way of building communities – they allow people to find others who share similar interests, passions and concerns. And like any communities these develop their own rites and traditions. A recent study of interactions on Twitter found a strong correlation between communities and the language they use – another example of social media mirroring behaviour offline.
Communities have always been defined by the language they use – as much to define what they discuss among each other as to exclude people who are not part of the group. An argot is a ‘secret language’ used by such groups which, at its simplest, just involves the groups using specific words and language. Words that define the community that uses them, but also excludes others from joining.
The research analysed 75 million Tweets between 189,000 users to identify the communities that exist and then looked at the words these communities used. The communities they found included the following (names my own):
- The SXSW-goers used words like ‘sxsw’, ‘tweetup’ and ‘conference’
- The educationalists used words like ‘edtech’, ‘pln’ and ‘edublogs’
- The Bieber Believers used words like ‘bieber’, ‘pleasee’ and ‘<33″
- The Twlighters used words like ‘kstew’, ‘robsessed’ and ‘twilight’
A full list of communities, and the words they use, is available in the original research.
But, perhaps more interesting that looking at how these communities can be defined by language is to look at how you can tell a community somebody should belong to by the words they use. Given a random sample of 1,000 words any given person uses on Twitter, it is possible to correctly say which community they belong to in 72% of cases.
This has potential benefits for anybody looking to identify or engage people in Twitter. By looking for patterns in the words that people use you can start to understand more about them, the issues they care about and the communities they are part of.