The Digital Pasts Lab (DigPasts-Lab) is geared towards performing innovative, quality research in digital humanities, with an emphasis on historical research
The Digital Pasts Lab rediscovers the past through innovative, quality research in digital humanities, especially in the fields of ancient history and Near Eastern texts. The lab is an interdisciplinary workspace between different humanities researchers and data scientists. Its overarching goal is to introduce machine learning and natural language processing to the study of history: we develop innovative computational methodologies to investigate primary sources left by past civilizations. Topics of research include machine identification and translation of cuneiform signs, human and machine cooperation, spatial data analysis, social networks, and linked open data.
The lab students and researchers participate in leading conferences and research networks for digital humanities (Pelagios, DHSI, ADHO, Linked Pasts, CHNT, Google Summer of Code). Scientific activities are supported by a variety of peer-reviewed research grants from the Israel Science Foundation (ISF) and the Israeli Ministry of Science & Technology (MOST). With their aid, new natural language processing models and optical character recognition methods for studying ancient cuneiform texts were and are in the process of development. There are currently close to a dozen ongoing projects led by the lab director, Dr Shai Gordin, with the participation of students and researchers from the MA to the Post-Doctoral level. Most of these projects have international collaborators from the fields of assyriology, archaeology, computer science and the digital humanities.
Dr Shai Gordin
Gordin’s cutting-edge research in the lab combines solid philological-historical approaches with modern tools and techniques of digital humanities and data science. In addition to publications in high-end interdisciplinary journals, Gordin published widely, in more than twenty-five peer reviewed venues, on the history and the textual sources of ancient Babylonia and Anatolia. For example, he investigated interactions of elite social groups and foreigners with the state, treated matters of historical geography, topics of social mobility, group interaction, community and state building, and identity.
Gordin’s main project in the lab, still ongoing, is the Babylonian Engine. Its purpose is to be a rich online portal for the study of cuneiform cultures. It includes special artificial intelligence models for cuneiform texts, such as one that restores broken segments using recurrent neural networks (DOI: 10.1073/pnas.2003794117), and one that transliterates and segments cuneiform signs to the Latin alphabet (DOI: 10.1371/journal.pone.0240511). In another project, Gordin created together with a PhD student a linked open data gazetteer of ancient place names in southern Babylonia. The code and data are freely available on GitHub and the tools are online for the scholarly community and public use (https://github.com/DigitalPasts).
The methods and projects devised by Gordin and the lab members, create a bridge between cuneiform studies and computer sciences. This bridge offers unprecedented opportunities for studying historical artefacts, textual or otherwise, and gaining new insight which was impossible beforehand. This bridge also brings public and scientific attention to the rich cultural heritage of ancient Iraq, the so-called “cradle of civilization”, as well as state-of-the-art preservation and digital curation methods for hundreds of thousands of its ancient artefacts.
