Amazon Translate is a neural machine translation service that offers fast, high-quality, affordable, and customizable language translation. It is important for machine translation to be accurate, fluent, and contextual when translating from one language to another. Many government and commercial organizations require domain-specific and language-specific customizable terminology.
Custom terminology allows you to customize your translation output by translating domain-specific vocabulary, such as brand names, character names, model names, and other unique content (named entities), exactly as needed. To use this feature, you need to create a terminology file (in CSV or TMX format) and specify it as a parameter in an Amazon Translate real-time translation or asynchronous batch processing request. You can refer to the “Customize Amazon Translate output to meet your domain and organization specific vocabulary” guide to get started with custom terminology.
In this post, we will explore the enhancements made to the custom terminology feature. These enhancements go beyond simple match and replace, and instead provide context-sensitive match and replace, which preserves the sentence structure. This enhancement aims to generate translations of higher quality and fluency by creating contextually appropriate versions of matching target terms.
To demonstrate the enhanced custom terminology features, we will use a custom terminology file. You can find instructions on creating a custom terminology in the “Customize Amazon Translate output to meet your domain and organization specific vocabulary” guide.
We will translate the sentence “she was a great tutor” using Amazon Translate. Here are the steps:
1. Go to the Amazon Translate console and choose “Real-time translation” in the navigation pane.
2. Select the “Text” tab.
3. Choose “French” as the target language.
4. Enter the text “she was a great tutor.”
The translation in French is “elle était une excellente tutrice.”
Now, under “Additional settings,” select “Custom terminology” and choose your custom terminology file.
The translation in French changes to “elle était une excellente éducatrice.”
In the custom terminology file, we specified the translation for “tutor” as “éducateur.” In English, “tutor” is gender neutral, but “éducateur” in French is masculine. Instead of performing a simple match and replace, the custom terminology feature used the target word and applied the correct gender based on the context.
Let’s test the feature with the source sentence “he has 10 sheep.” The translation in French is “il a 10 agneaux.” We provided custom terminology for “sheep” as “agneau,” which means “baby sheep” in French. The target word is changed to inflect plural.
Similarly, the source sentence “walking in the evening is precious to me” is translated to “me promener le soir est précieux pour moi.” The custom terminology target word “promenant” is changed to “promener” to inflect the correct verb tense.
The source sentence “I like burger” will be translated to “J’aime les sandwichs” to inflect the correct noun based on the context.
You can also test sentences with the target language as Spanish. For example, the source sentence “any action-specific parameters are listed in the topic for that action” is translated to “odos los parámetros especificos de acción aparecen en el tema de esa acción” to inflect the correct adjective.
However, it is important to note that some words may have different meanings based on context. For example, the word “order” in English can be a sequence or a command. Without explicit information, it is difficult to know the intended meaning. In this case, “order” should not be translated as “commande” in French.
In conclusion, the custom terminology feature in Amazon Translate allows you to customize translations based on your domain or language constructs. The recent enhancements to this feature create contextually appropriate versions of matching terms, resulting in translations of higher quality. Existing customers can start using the enhanced feature without any changes required.
For more information about Amazon Translate, you can visit the Amazon Translate resources page to find video resources and blog posts, and refer to the AWS Translate FAQs.