Ottoman Transliteration

Ottoman Turkish is a written language that was used in the Ottoman Empire from approximately the 13th century to the 20th century. Today, because the transition to the Latin (Roman) alphabet has been made and most of the words have fallen out of use, it is difficult to both read and understand Ottoman Turkish. A significant portion of the works that form the foundation of our national culture were written in Ottoman Turkish. Books, magazines, newspapers, notebooks, records, and documents in Ottoman archives and libraries hold an important place within the cultural, artistic, and historical heritage spanning hundreds of years. In order to quickly, effectively, and accurately access the information hidden in these sources, the help of technology, especially OCR, is needed. The TÜBİTAK-supported project "Artificial Intelligence-Assisted End-to-End Transfer from Ottoman Turkish to Contemporary Turkish" is a project we started for this purpose. The outputs of the project are presented to users on this site. With a holistic approach, the project aims to automatically translate Ottoman Turkish documents end-to-end, from document to Turkish text, in three steps: (i) Ottoman Turkish OCR, (ii) Alphabet conversion, and (iii) Language translation. Ottoman Turkish to Turkish Alphabet conversion can be defined as the transfer of a text from the source alphabet to the target alphabet. The purpose of this transfer can be either writing in the target language or reading in the target language. To avoid confusion, we refer to the transfer made for the purpose of writing in the target language as orthographic alphabet conversion, and the transfer made for the purpose of reading as phonetic alphabet conversion. In linguistics, orthographic alphabet conversion is known as letter conversion, letter transliteration, or transliteration, while phonetic alphabet conversion is known as transcription, phonetic transcription, or transcription. In alphabet conversion, the source and target alphabets are always different.