We're trying to train our own coreference model(s). The OpenNLP documentation points to org.annotation.opennlp.OpenNlpCoreferenceAnnotator
The main method requires a "data directory" and one or more text files. Could someone explain what they are and their formats? We don't have Treebank style parse in our data.
Sorry, I also never trained the coref component. Do you know if there is any freely available training data ?
I would like to improve the process and make it easier as I did for all the other components.
Log in to post a comment.