9/19    I work at an internet company and we have a bunch of data. I'm
        trying to find correlations with the data we collected and I'm
        wondering if applying Pearson's correlation coefficient to some
        data types is an effective way to approach this (since it's
        relatively easy to implement). Have you guys done something like
        this? I justs want something quick and easy and can be done in a
        day or less (e.g. I don't have time to build a full Bayesian
        classifier).