This research—funded by the Scientific and Technological Research Council of Turkey (TÜBİTAK)—examines the transformation of property regimes during the Ottoman–Greek border demarcation of the 1880s through a multilingual archival framework. Working with sources in Ottoman Turkish, English, French, and Greek, the project develops tailored approaches for each language and paleography type: a fully manual method for handwritten Ottoman documents, a semi-digital text-analysis strategy for English and French printed materials, and a deep learning pipeline for Greek newspapers.

The Greek component pilots a custom process combining automated image cropping, resolution enhancement, multilingual OCR, and lemmatization of property-related terminology using a four-language lexicon (Turkish, Greek, English, French). These steps generate high-quality textual data for term spotting and topic modeling. The pipeline integrates rule-based matching and vector-based semantic search, enhanced with large language models, to detect both known and previously unrecognized terminology across languages. Together, these multilingual and multi-method approaches establish a scalable model for historical text processing and cross-linguistic analysis of property and legal transformation in the late Ottoman borderlands.

Speakers: Fatma Öncel (Bahcesehir University), Doruk Şen (Istanbul Bilgi University), Ayşe Ozil (Sabanci University), Özge Sıla Çakmak (Istanbul Bilgi University)

Join us in person or online via Zoom.

Lunch will begin at 11:45 a.m. for in-person attendees.