Classifying 2 near identical documents

[Antonio_Campos1 (António Campos)]

Hello everyone.
I am conducting some tests on DU. I have two sets of electricity bills. One set shows a positive value like (total to pay: 25 um) with entity, reference, and that value for payment; another set, which is identical to the first, however, displays a negative value like total to pay: (- 2.5 um) without entity, reference, and value.
I ask:
How to use taxonomy so that when classifying, UiPath distinguishes these two types of documents?
Thank you.

[postwick (Paul)]

@Antonio_Campos1

with entity, reference

If one document has this and the other doesn’t, you should be able to use that to identify them. Have you trained the keyword classifier?

[Antonio_Campos1 (António Campos)]

Hello everyone. Thank you very much for answering my question. After searching, I found a way to determine the different documents. Thus:
1 - In the taxonomy, I added another type of document, taking care to point out the different fields compared to the two types of documents.
2 - In the keyword classifier, I changed the words for UiPath to take into account the differences. It already works, but this was only possible because I saw that at the time of word classification there are two possibilities for the same type of document:
In the first, if I put several words one after the other separated by a comma, then I am saying that I want to find them all in the same document (a kind of AND). On the other hand, by placing the words on each line, I get a procedure similar to OR, that is, just one needs to exist for the document to be classified with that type.
From the tests I did, it worked for the identification of two types of documents (which were almost identical at first).
Anyway, thank you very much for your help!