Popularity does not equate to universal authority

Popular Authority

Loïc Le Meur feels that Twitter needs to implement search by authority where authority is determined by the number of followers a user has:

We're not equal on Twitter, as we're not equal on blogs and on the web . . . Most people use Twitter with a few friends, but when someone who has thousands, if not tens of thousands of followers starts to speak, you have to pay attention.

I respectfully disagree with Loïc; both his assumption and his conclusion are overly-simplistic and do not take into account that popularity within a certain community does not equate to universal authority across all subject areas.

Take Loïc, for example. I respect his entrepreneurship and would definitely lend weight to his thoughts on online video or tech conferences. I'm sure the same is true for his 16,455 followers. I'm not entirely sure if I could say the same thing about his thoughts on stem cell research or religion. Should Loïc's tweets on religion feature higher than someone with fewer followers? What about his thoughts on brain surgery? Surely the thoughts of a brain surgeon with a hundred followers should rank at the top while Loïc's should rank pretty low on the list. (These are, of course, totally hypothetical — I don't believe I've seen Loïc tweet about any of these things.)

What Loïc's analysis is missing is the importance of semantics. And popularity does not contain any semantic value beyond telling us that the person is possibly interesting to a group of people for some reason. The pertinent questions that it does not answer are which group of people and for what reason?

Chris(tine) Crocker is hugely famous, with over 200,000 subscribers to his channel on YouTube. Were he to grace the side-streets of Twitter, he could probably muster a following list to easily eclipse any other name on Twitter. Do I want Chris Crocker's tweets at the top of my Twitter search results? Umm, not really. (Although I admit that I do occasionally watch his videos; I don't know whether it is for their train wreck value or in search of their hidden genius.)

That said, I'm not against the idea of understanding and associating authority on Twitter; both to people and to their tweets, but the solution is far more complex than a simple follower count.

To associate authority with a person on Twitter, you have to understand the topic(s) that that person is an authority in. A large number of followers may mean that they went on camera and cried out their eyes while yelling "leave Britney alone!" It does not necessarily mean that they are an authority in nuclear science even if their latest tweet contained the word "nuclear". This is a complex task and requires a web of authority similar to how a web of trust functions in security.

Associating authority with specific tweets might be easier to achieve. Robert Scoble provides a number of suggestions for associating authority with individual tweets, based on inspecting tweet metadata such as the number of times a tweet is favorited or retweeted. Even then, though, a purely numbers approach, without understanding the semantics, would not provide the expected result. If a tweet on a certain subject is so uninformed as to be laughable, it may get retweeted as a form of ridicule. Again, we need to understand why a tweet is retweeted in order to make assumptions on authority.

Ideally, determination of authority on Twitter search results would need to use a combination of personal authority based on a web of authority and content-metadata-derived authority that takes semantics of the message as well as the semantics of relationships into account.

But enough theory, following Loïc's impassioned call to arms, Jon Wheatley went and built an app called Twitority. So, does a purely numbers-based authority search really bring back trustworthy, informative results from authoritative individuals? No, it doesn't. Searching for "birth control" on Twitority, for example, brings back this gem from Twitter authority ImSleepDeprived as the first result:

I thought a profilactic [sic] was birth control. Isn't it??

Exactly the sort of though-provoking result we would expect from the person with the highest number of followers on Twitter who happened to mention birth control in a tweet.

Further reading: