Since November 2025, we have been publishing the proceedings of The Digital Orientalist’s Virtual Conference 2025 (AI and the Digital Humanities). While most speakers contributed to the written proceedings, four speakers chose to kindly share their work exclusively through recordings of their presentations, now available on our YouTube Channel. This post provides a short introduction to each of those contributions.
Keynote speaker and long-time team member at The DO, Ephrem A. Ishac (Austrian Academy of Sciences) presented “Who is the Public in a Digital Age? Democratization of HTR—Syriac as Example.” In the first half of his presentation, Ishac sought to define the public and their importance to work being undertaken in the digital humanities. In addition to this, he described different barriers to the democratization of HTR (Handwritten Text Recognition). Following this introduction, Ishac explored HTR as a force of democratization introducing the first public Syriac Handwritten Text Recognition (previously explored in The DO, here) as an example. He then turned to debates surrounding the costs and accessibility of different commercial HTR creation platforms. The final part of the presentation introduced various breakthroughs that could be achieved thanks to HTR and OCR in the world of Syriac Studies, before returning to the topic of “the public.” The full presentation can be viewed below.
Prince Kumar (Université de Franche-Comté) presented on the topic of “LLM-Assisted Geospatial Mapping of 17th-Century European Travel Writings on India.” This presentation provided a practical example of how LLMs (Large Language Models) can assist in geospatial mapping. He starts by introducing the background to his project and the various complexities that a digital humanist might face when exploring data from historical travelogues. Following this he provides a justification for using LLMs in this sort of project and compares “traditional” and LLM-centred workflows. Prince then introduces his preliminary findings and explores the limitations that he faced. His presentation can be viewed at the link below.
Tiziana Pasciuto (University of Turin), who joined The DO this academic year, presented “From Historical Encounters to Digital Records: The REDMIX Archive of Red Sea Mixedness.” Pasciuto introduced the REDMIX project—a planned, multilingual digital archive that documents the cultural, historical, and social mixedness of the Red Sea from the 1800s to the 2000s. She starts by introducing the key gaps that exist in current research, before exploring how the archive will be built, the digital tools that it will feature, and other factors that are helping to shape the platform. Following this, Pasciuto introduces the various challenges that creating such an archive entails and how these can be overcome. She closes by offering thoughts on future development. Please view the video below.
The final piece to be introduced today is Peter Francis Smith’s (University of Oxford) “Handwritten Text Recognition of Chinese Manuscripts: Complexities, Human Input, and Machine Learning.” Smith introduces his work on HTR for Chinese manuscripts. Following an introduction to the project background, Smith speaks about various steps of the HTR process including image selection, binarisation, segmentation, and transcription. As he covers these themes he introduces specific examples and challenges that arise in the use of Chinese manuscripts including the large number of characters, data loss through binarisation, dealing with complex layouts etc. Smith’s presentation can be viewed below.
