Text documents vertical search engine developed in Java, using Lucene library for indexing. The classification of documents by their topic was done using Logistic Regression algorithm. Performed regularization, polynomial and linear kernel, Frequency Based and Mutual Information feature selection.