is there a method to extract quote and their related speaker with the gestion of coreferecne
i want in output get a dic with [{"speaker" : , "quotes": }] and if we dont find the speaker , we put None to speaker and add "Potential Speaker" : coreference
If you want to identify a quote and the person who said it, you can use a model that has been trained with a database that contains quotes along with their respective speakers. Citron - https://github.com/bbc/citron is one such tool that you can use to label quotes. Because the model has already been trained, it can be used by you and other users.
I am unsure what domain your document pertains to - medical, IT, or something else? Therefore, my suggestion would be to fine-tune the pre-trained models offered by libraries like Citron to make them more specific to your domain.