Naive Bayes Classifier

Implement a simple Naive Bayes text classifier. Given training data as a list of (text, label) tuples, build a model and predict the label for a new text. Use word counts (bag of words) and Laplace smoothing.

Example

train = [("good great awesome", "pos"), ("bad terrible awful", "neg"), ("good awesome", "pos")]
predict = naive_bayes_predict(train, "good")
# => "pos"