Understanding Neural Machine Translation

Now that we have an appreciation for how machine translation has evolved over time, let's try to understand how state-of-the-art NMT works. First, we will take a look at the model architecture used by neural machine translators and then move on to understanding the actual training algorithm.

Intuition behind NMT

First, let's understand the intuition underlying an NMT system's design. Say, you are a fluent English and German speaker and were asked to translate the following sentence to English:

Ich ging nach Hause

This sentence translates to the following:

I went home

Although it might not have taken more than few seconds for a fluent person to translate this, there is a certain process involved in the translation. First, you read the German sentence, and then you create a thought or concept about what this sentence represents or implies. And finally, you translate the sentence to English. The same idea is used for building NMT systems (see Figure 10...