For the following sentence s expression result should be as following:
sentence:
Medical advice ranges from the confusing to the disapproving, while a financial disincentive could even transpire in the form of a long-mooted tax on soft drinks.
result:
(ROOT
(S
(NP (NNP Medical) (NN advice))
(VP (VBZ ranges)
(PP (IN from)
(NP (DT the)
(ADJP (JJ confusing)
(PP (TO to)
(NP (DT the) (NN disapproving))))))
(, ,)
(SBAR (IN while)
(S
(NP (DT a) (JJ financial) (NN disincentive))
(VP (MD could)
(ADVP (RB even))
(VP (VB transpire)
(PP (IN in)
(NP
(NP (DT the) (NN form))
(PP (IN of)
(NP
(NP (DT a) (JJ long-mooted) (NN tax))
(PP (IN on)
(NP (JJ soft) (NNS drinks))))))))))))
Your api response as there is only one node:
{
“text”: “(S\n (GPE Medical/JJ)\n advice/NN\n ranges/NNS\n from/IN\n the/DT\n confusing/NN\n to/TO\n the/DT\n disapproving/NN\n ,/,\n while/IN\n a/DT\n financial/JJ\n disincentive/JJ\n could/MD\n even/RB\n transpire/VB\n in/IN\n the/DT\n form/NN\n of/IN\n a/DT\n long/JJ\n -/:\n mooted/VBN\n tax/NN\n on/IN\n soft/JJ\n drinks/NNS\n ./.)”
}
ディスカッションに参加しましょう-以下にコメントを追加してください:
Ok, thanks.
Hi, the tagging API does not do deep parsing. It only does phrase level chunking, and by default, focuses on NE chunks. Unfortunately, while the demo supports choosing the tagger/chunker, and treebank or conll2000 will produce the results you want, I do not have that enabled right now for the tag API.