Fridy Mandita, Harnan Malik Abdullah, Toni Anwar, Panuwat Assawinjaiptech
The sign language research area shows many significant results in recent years. Mostly, the research of sign language is built for signer independent are focuses in sign and video corpus. Rarely, the research discusses about 'pola' grammar of sign language. The purpose of this paper is to contribute the topic of 'pola' grammar for sign language. The 'pola' grammar is used with a rule based algorithm and a Markov model and rule based. The model is used to validate the Malay sign language. The input of words and sentences of Malay sign language are taken from books, newspapers, and magazines by users either in form of single words or a paragraph of sentences. Training and testing words and sentences to check the 'pola' grammar of Malay sign language is implemented to gain the results of experiences. Simple sentences and complex sentences of Malay sign language are designed for a test case. The comparison of two models between rule based algorithm and a Markov model and rule based has been investigation based on the accuracy of tagset of Malay sign language. The accuracy of a Markov model and rule based shown a better result than rule based algorithm when used to check simple and complete sentences of Malay sign language. © 2018 IEEE.
Informatics Department, University of 17 Agustus 1945, Surabaya, Indonesia; Diploma in Computer Engineering, University of Brawijaya, Malang, Indonesia; Computer and Information Sciences, Universiti Teknologi Petronas, Perak, Malaysia; School of Information Computer, and Communication Technology (ICT), Sirindhorn International Institute of Technology (SIIT), Thammasat University, Bangkok, Thailand
