REVEALING THE TRUE AUTHORSHIP: A CASE STUDY

Abstract

As digital media is growing, the problem of text proliferation is becoming a big problem. Therefore, the identification of true ownership of a document has become a cumbersome task. In the digital era, it is very easy to copy someone's document and publish it in their name. So it is very necessary to check the true authorship. Authorship attribution becomes difficult when we conduct it manually. However, this process needs automation, when the document size becomes large. AA is a mixture of art, science and technology that helps to discover the genuine authorship of an unknown text/document, based on its specific writing features. These specific features can reflect the author's mood, education, gender, age, ideology, religion, or motivation. Many kinds of characteristics, like lexical, character, structural, syntactic, and semantic are used in authorship recognition. In this experiment, we used approximately 120 different kinds of feature sets. In our experiment, we examined that the logistic classifier was working well and gave good results in the form of accuracy.