Abstract— Javanese is a multi-level language; it comprises four speech levels used to convey local politeness. However, the negative tendency is detected regarding the use of Javanese speech levels among teenagers. They prefer to use the Indonesian national language (bahasa Indonesia, BI) because of the Javanese speech level complexity. A combination of statistical and memory-based machine translation is designed by the authors to help Javanese youths in translating between both languages. The evaluation shows that translating speech levels of Javanese to Indonesian is more accurate than translating in the opposite direction, as revealed by the average accuracy (A) of 0.83 for Javanese-Indonesian translation and 0.68 for the other direction.
Index Terms— Indonesian, Javanese speech levels, machine translation.
Aji P. Wibawa is with the School of Engineering, University of South Australia, Adelaide Australia and Department of Electrical Engineering, State University of Malang, Malang, Indonesia (e-mail: aji.wibawa@mymail.unisa.edu.au; ajipw@um.ac.id).
Andrew Nafalski is with the School of Engineering, University of South Australia, Adelaide Australia (e-mail: Andrew.nafalski@unisa.edu.au).
A. Effendi Kadarisman is with the English Department, State University of Malang, Malang, Indonesia (e-mail: effendi_kadarisman@gmail.com).
Wayan F. Mahmudy is with the Department of Computer Science, Brawijaya University, Malang, Indonesia (e-mail: wayanfm @ub.ac.id).
[PDF]
Cite: Aji P. Wibawa, Andrew Nafalski, A. Effendi Kadarisman, and Wayan F. Mahmudy, " Indonesian-to-Javanese Machine Translation," International Journal of Innovation, Management and Technology vol. 4, no. 4, pp. 451-454, 2013.