Word prediction algorithm in resolving ambiguity in Malay text

Word prediction algorithms are widely used in aiding disables by recommending the next potential word to be used. In this paper, the word prediction algorithm, n-gram, is used to resolve ambiguity in Malay text to retrieve relevant information. The test collection consists of Malay translated Al-Quran and Hadith. Initial experiment is conducted using the keyword madu that retrieved 72 Hadiths and 3 translated Al-Quran documents. Experiments are conducted on these retrieved documents using bigram and trigram. Result shows that trigram gives better prediction and retrieved relevant documents compared to the bigram. © 2016 IEEE.