What is it about?
In recent years, Neural Machine Translation (NMT) with attention mechanisms has emerged in research and industry. This study discusses the essentials of NMT (Seq2Seq, Convolutional Seq2Seq (ConvSeq2Seq), Recurrent Neural Networks (RNN), and Multi-Head Attention (MHA)) while implemented in formal passages in English-Indonesian and Indonesian-English. In general, the experimental results show the ConvSeq2Seq outperforms other methods.
Featured Image
Photo by Leonardo Toshiro Okubo on Unsplash
Why is it important?
This study explores four essential neural machine translations techniques: Sequential-to-Sequential (Seq2Seq), Convolutional Seq2Seq (ConvSeq2Seq), Recurrent Neural Network (RNN), and Multi-head Attention (MHA) while implemented in formal passages in English-Indonesian and Indonesian-English. This study also explains several fine-tuning concept strategies/architectures to improve NMT systems.
Perspectives
This study has been part of the author's independent research works series since 2014. This work adds perspectives to academic readers and society about the translation approach usually embedded in a search engine/recommendation/web mining system. The findings of this study give information on the significant characteristics of the four most popular neural machine translation techniques, showing that the right choice of algorithm and network architecture and fine-tuning on setup parameters have a considerable influence on its translation result.
Dr. Diyah Puspitaningrum
Read the Original
This page is a summary of: A Study of English-Indonesian Neural Machine Translation with Attention (Seq2Seq, ConvSeq2Seq, RNN, and MHA), September 2021, ACM (Association for Computing Machinery),
DOI: 10.1145/3479645.3479703.
You can read the full text:
Contributors
The following have contributed to this page







