Identify the Authors of Electronic Messages Through the Analysis of the Type and Style Based on Machine Learning Technique

Samira Zangoei,; Hassanali Nemati Shamsabad,

doi:10.35050/JIPM010.2014.042

Identify the Authors of Electronic Messages Through the Analysis of the Type and Style Based on Machine Learning Technique

Authors

Hassanali Nemati Shamsabad

10.35050/JIPM010.2014.042

Abstract

Identifying the author of an electroni message is one of
the main problems in text classification and natural language
processing. The aim of this article is to determine the authors of 50
cyber messages (by 50 potential customers, according to Amazon 's
website), by a machine learning methods. To evaluate the
effectiveness of the proposed method, the decision was carefully
tested and the results were compared with the performance of
machine learning methods. Also, when extracting various features of
authors' writing style for evaluation by machine, we tried to maximize
the features required to identify a writer. Therefore, nearly 10,000
different features were extracted from different entries in four
categories: lexical features, syntactic features, special features and
structural features. In this study, the average accuracy of the proposed
method reached to 98. 78.

Keywords

Identification of Authors

Machine Learning Methods

Characteristics of Writing Styles