Skip to content

Latest commit

 

History

History
44 lines (30 loc) · 2.62 KB

README.md

File metadata and controls

44 lines (30 loc) · 2.62 KB
Raw

app-cocohub

What is app-cocohub?

This is the project generator for COCOHub's crowdsourced translation service. It automates the creation of translation projects on COCOHub, and to do this it standardizes the data formats that a translation project uses. We use this in order to spend time only on the collection and verification of data, not on the minutiae of editing the same html templates to change a few words, a task that is rather easy to automate.

How do I use it?

To use this repository's tools effectively, you'll need to decide if you want to use the frontend+API, or the API alone. In both cases, you'll need an API key that you can obtain by going to your user settings at this url:

https://cocohub.cc/account/<your-username>/update and reset your API key. With that, you may continue to the next steps.

What you'll need

  1. A COCOHub API key (obtained above).

  2. You'll need a project.json file, containing the following keys: name, short_name, description.

    • name is the name of your project. An example is "Afrikaans Language Project", here: https://cocohub.cc/project/Afrikaans/
    • short_name is what will be used in the project's URL. For exmaple "Afrikaans" in the URL above.
    • description is a description of the project, i.e. what will be used in the project's overview.

  3. This step is where data is added to the project. You'll need CSV file containing the sentences you'd like to translate, and their identifiers which will help to keep track of the data generated by your community after your translation project is complete. To enable some automation of tasks for our projects, we need a consistent format for the colums in this CSV file, so we have set up the following convention for column names:

    • question: This is usually the text "What is the translation of this sentence?", but may be written in any way you like, including in the source language of your project, for cultural consistency.

    • id: This is a numeric identifier, unique to the project and may be parallel across projects within that specific COCOHub category i.e. within the "MS-COCO Crowd Translation" category, the same id in another language's project will point to the same sentence in the source language, which in the "MS-COCO Crowd Translation" case, is English.

    • caption: This is the text to be translated. Each row of your CSV will have a different sentence.

With these three things, we're on our way to creating your first translation project on cocohub.cc!

Using the provided scripts

TODO