This chapter addresses the challenge of restoring ancient inscriptions using Artificial Intelligence (AI) and highlights the crucial role played by the collaboration between epigraphers and computer scientists in advancing both fields. We delve into the importance of inscriptions and their digitization, emphasizing the pivotal role of data in AI applications. From a computational linguistics perspective, we formalize the problem as a Natural Language Processing (NLP) task. We present various AI methods designed to tackle the restoration of ancient inscriptions and highlight significant achievements in the field, along with their associated limitations. The analysis reveals that while numerous advancements have been made, there is still room for improvement, particularly in refining data quality and focusing on enhancing the explainability of predictions. In conclusion, this study underscores the importance of interdisciplinary collaboration for gaining insights into history, culture, and civilization.
AI for the Restoration of Ancient Inscriptions: A Computational Linguistics Perspective
Locaputo, Alessandro
Primo
;Portelli, Beatrice;Magnani, Stefano;Colombi, Emanuela;Serra, Giuseppe
2024-01-01
Abstract
This chapter addresses the challenge of restoring ancient inscriptions using Artificial Intelligence (AI) and highlights the crucial role played by the collaboration between epigraphers and computer scientists in advancing both fields. We delve into the importance of inscriptions and their digitization, emphasizing the pivotal role of data in AI applications. From a computational linguistics perspective, we formalize the problem as a Natural Language Processing (NLP) task. We present various AI methods designed to tackle the restoration of ancient inscriptions and highlight significant achievements in the field, along with their associated limitations. The analysis reveals that while numerous advancements have been made, there is still room for improvement, particularly in refining data quality and focusing on enhancing the explainability of predictions. In conclusion, this study underscores the importance of interdisciplinary collaboration for gaining insights into history, culture, and civilization.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.