we use the popular Reuters 21578 collection of documents as our training dataset which can obtain the dataset from this site: http://archive.ics.uci.edu/ml/machine-learning-databases/reuters21578-mld/. The collection has a total of 123 categories. We modeling Naïve Bayes Classifier according to the following steps: 1.Parse XML documents to extract topics and related content. 2. Tokenize the documents and stem. 3. Create our dictionary of all words (i.e., vocabulary) in the collection and obtain a inverse document frequency (IDF) for each term. 4. Vectorize documents using the TF-IDF scores 5. Train the NB classifier. 4. Classify HTML documents.