Unsupervised Author Identification and Characterization

Author identification is a hot topic, especially in the Internet age. Following our previous work in which we proposed a novel approach to this problem, based on relational representations that take into account the structure of sentences, here we present a tool that computes and visualizes a numerical and graphical characterization of the authors/texts based on several linguistic features. This tool, that extends a previous language analysis tool, is the ideal complement to the author identification technique, that is based on a clustering procedure whose outcomes (i.e., the authors’ models) are not human-readable. Both approaches are unsupervised, which allows them to tackle problems to which other state-of-the-art systems are not applicable.