Skip to main content

Creation of a project: Classification

From the header, click on Create new project.

Screenshot 1-1

1/ Project tab​

Fill in the required fields:

  • Customer
  • Project type
  • Project name
  • Project deadline
  • Project description

Non-mandatory fields:

  • Default markers: appears on the item later
  • to delete: indicates an item to delete
  • to check: indicates an item to check
  • to discuss: indicates an item to discuss

You can define access levels for the imported item :

  • the admin
  • the data scientist
  • the user

Screenshot 1-2

2/ Files tab​

Screenshot 1-3

the fields :

  • Config file: config file
  • Annotations file: file containing the annotations linked to the item
  • Predictions file: file containing the predictions linked to the item

Unlike annotations, predictions come from a pre-trained model. After the import, we can choose whether to keep them or not.

As for the mandatory field:

  • item file: this is a list that must be prepared in advance including the type of item and a unique id

3/ Labeling tab​

You can define what you want to do with the project.

  • Task Title: defines the name
  • Task Type: defines the type

Screenshot 1-4

You can add a new label by clicking on Add a new label.

The fields :

  • Label Code: must be unique
  • Label Name: does not necessarily have to have the same name as Label Code
  • Color: defines the color of the label
  • Hotkey: allows you to create a keyboard shortcut
  • Label Description: describes the label

Screenshot 1-5

caution

Once these labels have been defined, it is no longer possible to modify them later.

4/ Annotation Guide tab​

Here you can define annotation rules.

Once the specifications of a project have been filled in, you can click on Add a new task to start filling in the details of another project.

Screenshot 1-6

ex config.json


{

{

"tasks": [],

"name": "DEMO: Classification text 3 - emails",

"client": "LJN",

"type": "text",

"highlights": [],

"description": "Demo",

"admins": [
"admin@test.com"
],

"users": [
"user@test.com"
],

"dataScientists": [
"data@test.com"
],

"defaultTags": [],

"showPredictions": true,

"prefillPredictions": true,

"filterPredictionsMinimum": 0.4,

"deadline": "2023-04-30T13:57:20.355Z",

"entitiesRelationsGroup": []

}

}