My co-worker and I had a fascinating discussion with some digital humanities folks this week. We went to the last 2014 meeting of the Digital Humanities Research Network at UW Madison, and just sat around talking research pipelines and tools, and, because the research in question was about Twitter, digressed to tweeting habits and the relative publicity of information and conversation on Twitter.
From this conversation, I would like to propose the following tentative taxonomy of Twitter types, to be deduced from someone’s Twitter feed using the variables of 1) tweet frequency, and 2) tweet type (retweet, reply, use of hashtags). Popularity (measured in retweets, favorites and replies) could be another metric that I’ll leave for now.
The curator is a person who carefully manages their twitter stream. They may not tweet often, but when they do, the tweet is carefully composed and stands alone. If they retweet, there is a specifc reason/meaning for the retweet.
1) tweet frequency: daily to weekly. 2) tweet types: mostly original, occasional retweets. limited hashtag user.
The storyteller is a person who tweets constantly. They are either updating you about what they’re doing/reading/thinking about or actually tweeting their thoughts in the form of an extended essay where many tweets flow together.
1) tweet frequency: hourly. 2) tweet types: mix of retweets, replies, and original tweets. liberal hashtag user.
The informer doesn’t have a lot of their own to say, but are frequently retweeting other people’s articles, job postings, funny images, and more.
1) tweet frequency: varies. 2) tweet types: mostly retweets.
The networker may post their own tweets, may retweet, but mostly, they have questions they want answered, or answers they want heard. Their tweets may not appear in your twitter feed unless you’re friends with all the people they’re talking with too.
1) tweet frequency: hourly - daily. 2) tweet types: original and replies, retweets are mostly of replies.
The listener doesn’t really want to use Twitter as their medium of engagement; they primarily use it to keep abreast of what’s going on or as a placeholder online.
1) tweet frequency: monthly - yearly. 2) tweet types: any kind, but more likely to retweet.
It would be really sweet to make a visualization of these taxonomy variables, mapping tweet type on the xy-plane, frequency on the z-axis, and color-coded by some kind of popularity scoring metric.