SU@PAN'2015: Experiments in Author Verification

We describe the submission of the Soa University team for the Author Identication Task, part of the PAN 2015 Challenge. Given a small set of documents by a single person and a \questioned" docu- ment, possibly of a dierent genre and/or topic, the task is to determine whether the questioned document was written by the same person who wrote the known document set. This is a hard but realistic formulation of the task, also known as author verication . We experimented with an SVM classier using variety of features extracted from publicly available resources. Our solution was among the fastest, and running time was an ocial evaluation metric; however, our results were not so strong on AUC and C1.

[1]  Nick Cercone,et al.  N-GRAM-BASED AUTHOR PROFILES FOR , 2003 .

[2]  Ruslan V. Sharapov,et al.  Using of support vector machines for link spam detection , 2011, International Conference on Graphic and Image Processing.

[3]  Efstathios Stamatatos,et al.  Computer-Based Authorship Attribution Without Lexical Measures , 2001, Comput. Humanit..

[4]  Kalina Bontcheva,et al.  Text Processing with GATE , 2011 .

[5]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[6]  Kalina Bontcheva,et al.  SVM Based Learning System for Information Extraction , 2004, Deterministic and Statistical Methods in Machine Learning.

[7]  Efstathios Stamatatos,et al.  A survey of modern authorship attribution methods , 2009, J. Assoc. Inf. Sci. Technol..

[8]  Shlomo Argamon,et al.  Computational methods in authorship attribution , 2009, J. Assoc. Inf. Sci. Technol..