A Novel Approach for Authorship Verification

Abstract

Listen

Translate article

The Internet is increasing with huge amount of textual data. The crimes also increased in the Internet along with textual data. The authorship analysis is one important area attracted by the several researchers to reduce the problems raised through the text in the Internet. Authorship verification is one type of authorship analysis which is used to verify an author by checking whether the textual document is written by the disputed author or not. The accuracy of authorship verification majorly depends on the features that are used for distinguishing the style of writing followed in the documents. In the previous works of authorship verification, the researchers proposed various types of stylistic features to distinguish the authors writing style. The researchers analyzed that the performance of authorship verification was poor when the stylistic features were used alone in the experiment. In this work, a new approach is proposed for authorship verification where the content-based features were used in the experiment. The term importance is computed by using term weight measure, and these term weights were used to calculate the document weight. The document weights of training document and document weights of test documents were compared to verify the test document. The proposed approach accuracy is good when compared with state-of-the-art existing approaches for authorship verification.