Twitter: the "Ten Theses" gain empirical validation
Following widespread retweeting of my Ten Theses on Twitter, research emerged yesterday that seems to confirm one of them with stats. In the first ever quantitative analysis of twitter, analysing a month's complete usage data, Haewoon Kwak et al, of South Korea's KAIST, have come up with the following conclusions (big PDF):
- 85% of trending topics are headline or persistent news
- Twitter users are 4 degrees of separation from each other
- 35% of retweets occur in the first 10 minutes
- After the first hop, retweeting becomes almost instantaneous
- Any retweeted tweet reaches an average of 1,000 other users
I take this as provisional validation of thesis III and VIII for starters.
There's a slideshow and abstract of the research here. The researchers write a summary paragraph whose last phrase deserves a lot more thinking about:
"We have crawled the entire Twittersphere and obtained 41.7 million user proﬁles, 1.47 billion social relations, 4, 262 trending topics, and 106 million tweets. In its follower-following topology analysis we have found a non-power-law follower distribution, a
short effective diameter, and low reciprocity, which all mark a deviation from known characteristics of human social networks."