用haskell构建一个直方图，比使用python慢很多倍。

I was going to test naive bayes classification. One part of it was going to be building a histogram of the training data. The problem is, I am using a large training data, the haskell-cafe mailing list since a couple of years back, and there are over 20k files in the folder.I was going to test naive bayes classification.