Entrepreneur Abdullah Tarık Ömeroğlu and his team have developed an AI-powered program transforming how researchers, students and history enthusiasts engage with Ottoman Turkish documents. By transliterating millions of pages into the Latin alphabet, this innovation greatly enhances accessibility and understanding of this rich linguistic heritage.
Ottoman Turkish, used during the Ottoman Empire, is taught in various academic fields, including history, literature and Turkish studies, as well as in specialized courses. To preserve and understand this complex language, a group of young entrepreneurs created an AI-driven program transcribing original Ottoman texts. With features like transliteration and advanced word search capabilities, the program allows users to navigate and analyze vast collections of Ottoman Turkish materials, making the language more approachable.
The project was born in 2020 when Ömeroğlu was a student at Boğaziçi University. While preparing books in Ottoman Turkish for a publishing house, he and his programmer roommate envisioned an AI solution to simplify access to these texts. After receiving support from TÜBİTAK BİGG (Individual Young Entrepreneurs Program), they formalized their idea and established their company at ITU Teknopark.
After three and a half years of development, the team successfully implemented AI models that categorize their system into individual users and libraries. Creating a pooling system for libraries hesitant to share their collections enables users to search through 7 million pages of documents, with translations completed in under a minute and a remarkably high accuracy rate.
Looking ahead, the team plans to introduce simplified Turkish versions of texts alongside chatbots and summarization systems that enhance user interaction and understanding. This next phase will further empower users to explore the context and themes of documents, offering answers to questions about authorship, dates and main ideas.
Ömeroğlu emphasized that their aim is not to profit from individual users, which is why they maintain minimal fees. Their comprehensive system integrates transliteration and extensive data searches, streamlining the research process for anyone interested in Ottoman Turkish.
Highlighting the contributions of collectors Hakkı Tarık Us and Seyfettin Özege, Ömeroğlu noted that the team collaborates with various libraries and collectors to expand their resource database continually. This commitment ensures a rich and growing collection, making Ottoman Turkish texts more accessible than ever.
