Multi-Class Text Classification w/o model in the loop

Apologies if this is answered somewhere else, but I have outputs from a posting summarization project, and I want to get a baseline for how well it's doing. Ideally, I want to display ten of the "topics" for each bit of text simultaneously, and be able to accept or reject each one within the same screen. How can I accomplish this? I have a recipe set up for accepting/rejecting one "topic," but not sure how to add others.

Thanks so much!

Hi! I think what you're looking for is the textcat.manual workflow – you pass in a source and a bunch of options and just label everything by hand as it comes in.

(Under the hood, it works pretty much exactly like this custom recipe example – so if you want to incorporate your own logic in Python, you could also write your own recipe with the choice interface.)