Davey Winder
Tuesday, 01 July 2008 07:07
Your IT -
Home IT
Page 2 of 3
It is other users, other Diggers, who are at the heart of
the concept. You see the Recommendation Engine tracks those users who
have Dugg a story before you and suggests that you might like to see
the other stories that they have liked as well. This all being based
upon your last 30 days of Digging history.
A very simple idea, but one with some merit. It
will be something of a learning process I suspect, but the more content
you Digg so the more data there is to analyse and the smarter the
Recommendation Engine becomes. Every time you Digg a story it will
match you with all the other Diggers who have done likewise. More
importantly, it keeps track of the Diggs that you have in common.
"When it's time to calculate your Recommendations, the Engine draws
from this pool of matched Diggers. For each matched Digger, it computes
a correlation coefficient between you and them. It then picks a cutoff
for this correlation coefficient, and the Diggers who make the cut are
called "Diggers Like You" explains Kast.
Understanding how these correlations are calculated is not difficult.
The Recommendation Engine simply takes the number of Diggs in common
between two users and divides it by your total number of stories Dugg.
The ratio being a correlation coefficient, a number that
reflects if you and the other user never or always agreed on Diggs.
If another user Diggs a lot more than you, then they must agree with most of your stories for them to become a Digger Like You.
Read on to discover more about the inner workings of the Digg Recommendation Engine...
CONTINUED
LATEST HEADLINES FROM YOUR IT
