A classification links an existing Machine Learning class to an element, with a confidence score.
It can also be used for
manual classification by human annotators.
Pending, when created by a workflow,
Validatedwhen approved by a human annotator,
Rejectedwhen invalidated by a human annotator.
confidencescore, set between 0.0 and 1.0 (percentage). It defaults to 1.0 when the classification is created by a human (we always trust human annotators)
Trueon the classification, when it judges that the confidence score is high enough (that score would be dependent on tools, models and context).
Classes and classifications can be viewed and modified on multiple pages of the web interface.
When browsing elements, you can apply classification filters using the filter bar:
When any of those filters are combined, then only elements that have a classification that matches all the specified criteria will be selected. An element with multiple classifications that match only one of each criteria would not be displayed.
When viewing the details of a single element, the panel on the right side has a Classifications section where you can:
You cannot create a new Machine Learning class from this interface, but you can apply an existing one (provided it's not already applied).
If a Machine Learning worker produces a classification with a high confidence (known high score), a gold medal icon will appear next to the class name:
Classes can be managed by project administrators by using the Manage classes action under the Actions menu in a project.
From the class management page, it is possible to add, rename and delete classes. Classes can only be deleted if they are not used by any classification on any element.
These endpoints are the most useful to handle Element classifications:
In order to retrieve the classifications within a list of elements, you can use
with_classes=true query paramenter on these endpoints: