twitter authority

this post got me thinking about twitter authority. here are my preliminary thoughts below.

first a few assumptions/facts:
#1 there are different types of users on twitter: the expert, the intensive user, the passive user, the creator, the focused user, the eclectic user…
#2 each user uses twitter differently: posting links, posting a short thought, conversing, listening, asking questions, retweeting…
#3 the power of the network lies in part on the ability for one “lone” voice to broadcast an important tweet and for all others to potentially find it, in other words the network is as strong and useful as its weakest link in a certain respect, all other aspects being equal.
#4 as it stands, using twitter is like drinking from a fire hydrant with little in the way of modulating the water flow.

the combination of #1 and #2 ensures that one user “profile” or “pattern” of use can be a good indicator of who one is: an influencer, or a connector, or an amplifier… in his/her respective “network” within twitter. it follows that there are many ways of evaluating or ranking a user.

#4 underlines the absolute necessity of organizing the data flow to eliminate noise, reduce the time it takes to find the tweets one wants to read, the users one wants to continue or start to follow. twitterdeck or stocktwits answer that need. i am sure more will come as twitter grows. other ways or organizing one’s twitter universe are being being devised and shared daily, see here.

so now comes the concept of ranking users quantitatively which i think is an excellent idea. actually twitter grade, which i have been playing with recently, does provide some sort of quantitative grading of users already, although outside of twitter itself. a first step would be to integrate twitter grade or a service based on a similar algorithm to twitter directly. a first step but not the only step. what one gets in visibility on a user’s following reach, one looses on the quality of a user’s tweets, hence potentially violating #3 above.

i would like to see a combination of quantitative authority ranking (quaar), qualitative authority ranking (qualar), dynamic authority ranking (dar) and semantic search filtering (ssf). let me explain myself:

- quaar: would provide a combination of basic quantitative metrics such as
a) # following overall, # following in geographical region giving an indication of “volume” or reach (and the derivative thereof, as in # following the followers…)
b) same as above but for # followed, giving an indication of “listening skills”
c) # of updates
d) # of times a user rt’s
e) # of times a user posts a link
f) mathematical combination of a), b) and c) giving a power quaar ranking ( apologies to twitter grader which is already providing such a service)

one’s profile would give these indications while only f) would appear on any given tweet for example.

- qualar: would provide a combination of basic qualitative metrics such as
a) # of tweets favorited giving an indication of breadth of “attractiveness” (or some other type of ranking mechanism at the tweet level)
b) # of times a tweet is favorited giving an indication of depth of “attractiveness” (or some other type of ranking mechanism at the tweet level)
c) combination of both a) and b) giving an indication of breadth and depth at once
d) # of rt giving an absolute indication of follower power
e) # of rt compared to followers for example, as a ratio, giving a relative ranking of follower power
f) # of links to a particular tweet
g) mathematical combination of the above giving an indication of power qualar ranking

b) and f) or g) could appear only on a given tweet, the rest on the profile for example

- dar: would provide a snapshot of the health or trending of a user or tweet
a) velocity of rt, links, favorite rating after a tweet giving indication of “immediate power” ranking
b) residual length of rt, links, favorite rating after a tweet giving indication of “long tail” ranking
c) mathematical combination of a) and b) giving indication of ambient awareness ranking
d) mathematical representation of how # following, # followed (and derivative thereof… # following the followers and so on and so forth) behaves over time, how fast up or down, giving indication of general health of a user’s network

a) and b) or c) may appear on a given tweet, the rest on a user profile

- ssf: some type of semantic search technology that would allow for a quick filtering and weeding out of low relevancy tweets, thereby focusing only on high relevancy ones.

ok, i think i will stop here, realizing fully that i did go overboard on a subject i do not totally master. one last thought though. devising some sort of authority ranking delivers a key to building and seeking relevance on twitter. relevance delivers in turn value to those active on the platform. increasing the value one derives from using twitter is the ultimate goal.

ok, i lied, one last more thought… the implicit given is that one would already have filtered the users one knows or one knows why one follows them as these users will presumably bring higher relevancy compared to unknown users. the above musings on ranking and authority only applying to the users one does not know.

:: :: :: :: :: :: :: :: :: :: ::