Applying Naive Bayes Classifier to Document Clustering

Document clustering partitions sets of unlabeled documents so that documents in clusters share common concepts. A Naive Bayes Classifier (BC) is a simple probabilistic classifier based on applying Bayes' theorem with strong (naive) independence assumptions. BC requires a small amount of training data to estimate parameters required for classification. Since training data must be labeled, we propose an Iterative Bayes Clustering (IBC) algorithm. To improve IBC performance, we propose combining IBC with Comparative Advantage-based (CA) initialization method. Experimental results show that our proposal improves performance significantly over classical clustering methods.