A Signing Avatar System for Turkish to Turkish Sign Language Machine Translation

(TÜBİTAK 1003, 15.10.2014 - 15.10.2016)

In our country, deaf students have great difficulty understanding written Turkish used in the MEB (Ministry of Education) curriculum, and cannot follow the course contents. When these students reach their fourth year at primary school, they can not compete even with the first year hearing students in terms of writing and understanding the written material. Some of these students try to make up for this by attending extra courses outside the school; however, since there is not enough  supporting educative material, they still face difficulties, and most of them prematurely finalize their education. This is an important problem not only for deaf students and their families but also for the hearing students and teachers in the integration classes (classes for both hearing and deaf students). The “Report on the Evaluation of Integrated Education Applications in Primary Schools” published in 2010 (MEB, 2010) presented the problems experienced in these classes. The aim of our project is to develop a computer-supported interface which is compatible with the technological infrastructure of Fatih project, and which will translate the curriculum into a form that deaf students can understand.  To summarize briefly, a novel system that will translate written Turkish into written TID will be designed and implemented. This system will also translate written TID into visual TID through animation / virtual characters. Such a study has not yet been done for TID. To achieve this, knowledge of grammar of written Turkish and TİD are required. However, the grammar of TID has not been fully understood and formalized yet. Therefore, the project will be carried out by analyzing the most frequently used structures in the selected corpus. Turkish, as a morphologically rich language with flexible word order presents challenges for natural language processing that are different from other widely studied languages such as English. Therefore, one cannot directly apply the methods and findings from other languages to Turkish. Our study will, thus, have a unique contribution also for languages such as Finnish, Hungarian and Korean which have linguistic properties similar to Turkish. This project aims at filling a crucial gap in the education of deaf students. 

In every country, different native sign languages are used (e.g. TİD in Turkey, ASL in the U.S.A. etc.), and these sign languages have linguistic properties that are different from the linguistic properties of the spoken language(s) spoken in those countries. There are studies that have been conducted in order to produce translation systems for deaf people to translate official documents and education materials from written text to sign language. There are active studies conducted on developing signing avatars for Tunisian, German, English, Dutch, French and American sign languages. However, sign languages are substantially different from each other as spoken languages are. Therefore, a system that has been developed for another language cannot be used for a Turkish-TİD translation system. A successful machine translation system has to be developed based on TİD grammar. In a few recent translation studies on Turkish, only selected words are translated into TİD signs (with pictures/photos, videos and avatar animations). However, this approach leads to incorrect translations at the sentence level. Turkish and TİD are different languages and as all spoken and sign language translation systems, this issue is a machine translation problem that has to be studied on syntactic and semantic levels. Because of this, it contains all the challenges of machine translation. Again because of this, the recent systems that have been developed for Turkish have been nothing more than a limited dictionary-like system that translates words to signs. To conduct research in this area, team work which combines the expertise of researchers in the fields of natural language processing, linguistic analysis of TİD and human-computer interaction is required. In our project, three different research teams from İTÜ and Boğaziçi University, with expertise in these fields come together. This team has been conducting a number of national and international projects that provide the background to the proposed project. In the project we propose, for the first time in Turkey, a real automatic machine translation system for Turkish-TID will be developed as a result of the combination of expertise of these teams.

​Project Team
Assoc. Prof. Dr. Hatice Köse (PI)
Assist. Prof. Dr. Gülşen Eryiğit (Istanbul Technical University)
Assoc. Prof. Dr. Meltem Kelepir (Researcher, Boğaziçi University)