Authorship Verification: An Approach based on Random Forest: Notebook for PAN at CLEF 2015.

Abstract

Authorship attribution, being an important problem in many areas including information retrieval, computational linguistics, law and journalism etc., has been identified as a subject of increasingly research interest in the recent years. In case of Author Identification task in PAN at CLEF 2015, the main focus was given on cross-genre and cross-topic author verification tasks. We have used several word-based and style-based features to identify the differences between the known and unknown problems of one given set and label the unknown ones accordingly using a Random Forest based classifier.