Year: 2024
Type: article
Abstract: Bahasa Melayu or Malay, is the national language of Malaysia. It is the country's official language and is widely spoken and understood by the Malaysian population. Although Bahasa Melayu is widely spoken across Malaysia, different dialects are spoken in different states. This may cause tourists in Malaysia to face language barriers due to various dialects spoken by local people. The Bahasa Melayu Dialect Translator and Detector system or MalayFY, is an innovative solution to overcome language barriers and facilitate effective communication between users of different dialects in Bahasa Melayu. This system consists of three modules: a language translator from English to Bahasa Melayu, the Bahasa Melayu dialect translator and the Bahasa Melayu dialect detector. The project's primary goal is to create a system that is capable of translating English text into Bahasa Melayu, translating standard Bahasa Melayu into dialects and detecting an appropriate dialect for a dialect sentence. The dialect detection module employs the Random Forest algorithm to predict the dialect of user input sentences. Nine dialects are covered in this system: Johor, Kedah, Kelantan, Melaka, Negeri Sembilan, Pahang, Penang, Perak and Sabah. The Bahasa Melayu dialect translator module is to translate Bahasa Melayu sentences to a specific chosen dialect. This module uses a simple rule that applies predefined linguistic rules and patterns to determine the appropriate translation. The evaluation results for the dialect detector and translator are promising, although with limited data available, especially for developing the dialect detector. The accuracy result of the dialect detector is 62%, which could be improved when more data are trained. Overall, MalayFY is successful in helping people who have problems understanding the Bahasa Melayu dialect, especially tourists.
Authors: Adella Java Dirgantari, Halizah Basiron, Nur Zareen Zulkifli, Fitrah Rumaisa
Institutions: Universiti Teknikal Malaysia Melaka, Universitas Widyatama
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4408312358