I have a selection of texts. Missed from via doc2vec from the gensim library. The result is good. Similar texts defines with a bang. How can you cluster texts?

I tried to do this: I got a vector for each text. Threw all this in k-means . The result was not very good.

What other approaches can be used with a trained doc2vec model?

  • "The result was not very good." - would indicate whether it was - what exactly turned out to be bad (and on what data with what parameters). Although, probably - I do not understand something because of amateurishness :-) ps however, I am a necropolis :-) - Alexander Pozharskii

0