[:es]Idatha seeks to transcribe with Artificial Intelligence manuscripts of José Enrique Rodó found in the National Library.

The project has been underway for about a month, and is in a proof of concept stage, seeking to build an AI model based on Computer Vision that facilitates the digitization of the Uruguayan writer’s notes.

Manuel Reynaert, Team Leader of the Idatha team, stated that if the realization of such a challenging project can be verified, «it would save years of manual work». He added that at the moment, the team members are using Python programming language with libraries such as Tensorflow, Keras and Pillow.

When asked about the progress that has been made, Manuel said that a month after starting, a new Idatha member, Zuhair Abbas, recently joined the project and that the first results will be obtained by the end of November.

José Enrique Rodó (1871-1917) is one of the greatest representatives of the generation of the 900 in Uruguay, he was an essayist, literary critic, politician and journalist, in 1900 he published Ariel, a work that was an immediate success, considered as the most notable work of the writer with great influence in Latin American culture and politics of the time. A multifaceted figure of Uruguay’s cultural heritage, his records formed the Rodó Archive, founder of the Literary Archive of the National Library, configuring the most extensive production of the collection with more than twenty-one thousand documents. By transcribing the author’s manuscripts with Artificial Intelligence, we seek to digitize and give a new life to a material that today remains stored and archived without seeing the light of day.

Learn more about this project by clicking here.[:]