Author Identification System

Abstract

Author Identification System is a desktop based application which uses two different machine learning algorithms to perform text classification on the articles written by different authors, hence identifying the correct author of an article. The system is first trained using a training dataset. The system calculates values of certain features such as number of sentences, average number of words etc. of an author and then uses the features to identify top three best matches for an article (unknown author) given as input to the system. In text classification language, AIS successfully performs Authorship Attribution. Author Identification System can be used in the field of digital text forensics and in the case of increasing number of plagiarism problems, forgery of text, or fake text. This system can contribute to play a key role to solve many different crimes, security issues, and civil transactions. The system was developed after researching the features and similar systems. It demonstrates the working of multiple algorithms to achieve text classification in the case of Author Attribution.