Authorship Identification Using a Reduced Set of Linguistic Features

The proposed solution for authorship attribution combines a couple of the most important features identified in previous research in this domain with classification algorithms in order to detect the correct author. We consider that the most relevant aspect of our work is the small number of linguistic features and the use of the same framework to solve both the open and the closed class authorship problem, by only changing the classification algorithm. This approach obtained an overall 77% accuracy with regard to the total number of correctly classified documents.