Text Classification from Labeled and Unlabeled Documents using EM
Distribution of the number of citations over years.