Neither the and/CC nor the not/RB were included in or made into chunks. Is this correct? That these two things don't belong in a chunk, or is it an error.
Thanks,
Theresa
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
That looks correct to me although I don't know the guidlines exactly. The "and" is part of a VP coordination and the not is also a VP modifier which I believe are typically left un-chunked. There is more information about the specifications of the data used to train the chunker at:
Hi -
When the chunker is given the following sentence as input:
washington/NNP says/VBZ no/DT ,/, because/IN the/DT taliban/NNP had/VBD no/DT clear/JJ chain/NN of/IN command/NN and/CC was/VBD not/RB a/DT legitimate/JJ government/NN ./.
it produces the following output:
[NP washington/NNP ] [VP says/VBZ ] [NP no/DT ] ,/, [SBAR because/IN ] [NP the/DT taliban/NNP ] [VP had/VBD ] [NP no/DT clear/JJ chain/NN ] [PP of/IN ] [NP command/NN ] and/CC [VP was/VBD ] not/RB [NP a/DT legitimate/JJ gover
nment/NN ] ./.
Neither the and/CC nor the not/RB were included in or made into chunks. Is this correct? That these two things don't belong in a chunk, or is it an error.
Thanks,
Theresa
Hi,
That looks correct to me although I don't know the guidlines exactly. The "and" is part of a VP coordination and the not is also a VP modifier which I believe are typically left un-chunked. There is more information about the specifications of the data used to train the chunker at:
http://cnts.uia.ac.be/conll2000/chunking/
I think the definitive guide however is in the perl code used to extract this information from Penn Treebank parses. Hope this helps...Tom