IMDB. COM ranking algorithm (Bayesian formula)

Because recently do the World wide brand word-of-mouth project, need to make a contrast to the word, now the library also has a part of the data, there are a lot of brand score incredibly is the same, this is the real data in the library, if I simply by the average score, also not fair, because there are many comments, but there are few.

So I quoted the Imdb.com ranking algorithm, he is mainly on the ranking of top250, the number of ratings have a certain limit, and our brand library always have to let the corresponding brand exposed. So I applied the Bayesian formula to this part of the data.

The final result is still very good, can achieve their desired effect. I feel a little embarrassed if I use the average.

Here to share with you the ranking algorithm formula:

WR = (V÷ (v+m)) XR + (M÷ (v+m)) XC

Specific meaning:

R = Average score for a single movie
v = number of effective ratings for a single movie
m = minimum number of effective ratings required to be selected for the top250 list
C = Average score for all films

IMDB. COM ranking algorithm (Bayesian formula)