Creating a corpus
June 1, 2020
A corpus contains sentence pairs that are used to adapt machine translations.
To create a corpus, go to the project's SOURCE page. Under the CORPUS tab, click Create corpus.
A card will appear on the screen. In the card, enter a name for the corpus.
Upload a CSV file to import sentence pairs into the corpus (ensure the file is formatted correctly). Please note that this step is optional.
Once the information for the new corpus has been input, click Create.
There are two ways to add sentence pairs to a corpus.
The first way is to import sentence pairs from a CSV file. In the corpus, click Import pairs.
A card will appear on the screen. In the card, upload a CSV file containing sentence pairs (ensure the file is formatted correctly).
Once the CSV file has been uploaded, click Preview & import.
The corpus will show a preview of the sentence pairs contained in the CSV file. Once the sentence pairs have been previewed, click Import (X) pairs.
The second way to add sentence pairs to a corpus is to manually create matching pairs in the source and target language.
To manually add sentence pairs, click Add pairs.
A popup will appear on the screen. In the popup, enter the desired phrase in the source language and target language.
Once the phrases have been input, click Add (X).
For additional instructions, please contact your AICX manager.