Join us on October 22nd, Tuesday, at 12 pm for a research lunch seminar with Merve Tekgürler, PhD Candidate in History and MS Student in Symbolic Systems at Stanford. They will talk about the use of artificial intelligence, specifically neural machine translation (NMT) and large language models (LLMs), to translate Ottoman Turkish into English. Their study consists of two approaches to bridge this lack of resources for computational tools for Ottoman Turkish. First, making use of tools developed for Modern Turkish, they fine-tune NMT models for Ottoman Turkish. Second, they test the translation performance of LLMs like Google’s Gemini and OpenAI’s GPT-4. While LLMs outperformed NMT in translation accuracy, ethical concerns arose due to content moderation filters that blocked parts of historical texts, such as depictions of violence. Their research highlights both the potential and limitations of AI for translating low-resourced languages. Join us for a talk on historical languages and early modern history, and implications in using AI to explore and analyze these histories and languages.

Lunch will be served. Zoom option also available.  RSVP here for lunch and Zoom information.

This event is co-hosted by Center for Spatial and Textual Analysis (CESTA), and Ottoman Empire and Middle East Workshop within the History Department. It is co-sponsored by the Ottoman-Turkey Encounters at Stanford.

About the Speaker

Merve Tekgürler is a PhD candidate in History (ABD) and an M.S. student in Symbolic Systems. They have a BA degree in History and Social and Cultural Anthropology from Freie Universität Berlin and an MA in History from Stanford. Merve’s dissertation, tentatively titled “Crucible of Empire: Danubian Borderlands and the Making of Ottoman Administrative Mentalities,” focuses on the Ottoman-Polish borderlands in the long 18th century (1760s-1820s), examining the changes and continuities north of the Danube River in relation to Russian and Austrian expansions. They study the Ottoman news and information networks in this region and their impact on production and mobilization of imperial knowledge.