In the context of the affirmation of Digital Humanities as an independent field and the subsequent application of computational methods to the analysis of literary texts, the project deviates from the widespread approach based on the representation of Authorial variants rather than on the much less practiced study of them.
The research purpose is to start from a critical apparatus of a manuscript to create an automatic system of detection for pre-established categories of Authorial corrections. Giacomo Leopardi’s Canti have been chosen as case-study for this work for Leopardi's manuscripts represent the testing ground for the development of Authorial Philology.
As a result, these newer techniques deepen the conventional philological reading, making it possible to gain linguistic and empirical insight into the creative process of an Author directly from the PDF version of the work. Such immediacy of the model will allow to widen the study to other Authors in order to make a complete analysis of the types of Authorial corrections and compare these different conceptions of writing statistically.