Corpus-based Sequence to Sequence with Luong Attention Mechanism for Ilocano to English Translation

Effective communication is a vital means of conveying ideas, yet language barriers often pose significant challenges for individuals and groups conversing in different languages and dialects. These challenges can result in misunderstandings and misinterpretations due to a lack of clarity in expression. Current Ilocano-English translation systems provide some assistance. However, they struggle with translating Ilocano accurately, especially when the input includes sentences with a higher number of words. To address this issue, our study developed a translator using the Seq2Seq model with Luong Attention Mechanism and employs Neural Machine Translation techniques. The developed translator application achieved notable results with an accuracy of 43.95% for English to Ilocano translation and 81.25% accuracy for Ilocano to English translation, along with a BLEU score of 2.50 and 15.42 across all n-gram levels (1-4 words), respectively. These findings contribute to bridging language barriers and enhancing cross-linguistic communication.