What are categories and text inside the classifier?

The classifier works with two elements: categories and text.

A category is a tag that the text should be assigned to. Text contains the content that should belong to a specific category.

For example:The following text should be assigned to the category «Change of address»:

To train your classifier properly, a minimum of 10 examples per category have to be uploaded. To get a qualitatively high result, we recommend significantly more examples. 

The principle that applies here is: The more examples you upload, the more accurate the classifier will be. Training examples can be added via data import or manually.