This document discusses web page classification using naive Bayes classifiers. It outlines the goals of web page classification, including improving web directories and search results. The document reviews literature on different representations for classification, including bags of words, n-grams, using HTML structure, and visual analysis. It then describes experiments using a university web page dataset to classify pages into categories like course, department, etc. using bag of words, HTML tag weighting, and n-grams. The document concludes with an overview of evaluation techniques like k-fold cross validation and confusion matrices.