Analysis with doc2vec learner and only one label

I’m playing around with the doc2vec learner. My workflow looks similar to the sentiment analysis workflow example. However there’s one big difference: my input strings have only one label: ‘good’. I’ve build a column of strings with words belonging together and labeled ‘good’. I want to test new strings with the Word Vector Apply node and label new strings with ‘good’ (strings that apply to my vector model; less deviation) and ‘bad’ (strings that do not apply to my vector model; more deviation). Is this possible?

Hi @marcv,

could you maybe attach an example workflow that shows what you are trying to do?