In Codice Ratio: Machine learning meets manuscripts from the Vatican Apostolic Archives
Dienstag, 23. November 2021
18:15 - 19:45 Uhr
Universität Zürich - Aula der alten Kantonsschule (Rämistrasse 59, Raum RAA G01)
Lecture by Serena Ammirati and Paolo Merialdo (Università Roma Tre)
In Codice Ratio (ICR) is a research project born in 2016 at Roma Tre University with the aim of creating a tool for text transcription, extraction and use of data from ancient and modern historical manuscript sources, using machine learning and crowdsourcing techniques.
Since its conception, the project has benefited from the collaboration of the Vatican Apostolic Archive (VAA): the starting point was the study and analysis of papal chancery registers from the 12th and 13th centuries, which were interesting both in terms of graphics and text. ICR sees the collaboration of software engineers, archivists and paleographers: the machine learning algorithms for the Handwritten Text Recognition (HTR) engine are trained by data that can be collected involving non-expert users by means of a user-friendly Web interface. In this way, we aim at making available to the scholarly community not only transcriptions of manuscript sources with a satisfactory degree of reliability, but also a search engine for scholars, who will have the chance to search documents through in order to find prosopographical datas, words, facts, dates, documentary formulas. A big portion of European medieval, modern and contemporary history can be investigated by exploring the documents kept in the VAA, documents that are nowadays not always easy to be accessed and read. Automatic transcription and search engine will make easier to access this evidence. Automatic transcription can also be intended as important didactic tools in teaching palaeography and history of the handwritten documents. In this communication we would like to tell you about our research experience, the results achieved so far, and the future prospects in the field of humanities research and beyond.
Diese Veranstaltung ist Teil der Vorlesungsreihe «Bücher, Daten, Räume. Die Hochschulbibliothek im 21. Jahrhundert»
Für diese Veranstaltung benötigen Sie ein Covid-Zertifikat. Bitte weisen Sie es beim Eintritt vor.