Hi all, I wonder if it's possible to train sentence detector to handle direct speech. Examples:
"Tonight," I said, "is Halloween!" "I'm hungry," he said. "Come here!" he said. "Where are you?" he asked. I'd like to analyze news articles and they are full of quoted citations. My questions are: - How to approach the problem? - How should be example sentence above entered into training file (one line for the whole sent.)? - Is it even possible to use current implementation of SentenceDetector for this? Regards, Tomas Blesa S pozdravem Tomáš Bleša ANNECA s.r.o.