def demo ():
# load train corpus
train_sents = brown.tagged('a')[:500]
# create taggers
tagger = MarshalNgram(3)
#tagger.train(train_sents)
#tagger.marshal("ngram.test")
tagger.unmarshal("ngram.test")
print tagger._model
def demo():
# load train corpus
train_sents = brown.tagged('a')[:500]
# create taggers
tagger = MarshalNgram(3)
#tagger.train(train_sents)
#tagger.marshal("ngram.test")
tagger.unmarshal("ngram.test")
print tagger._model
def pos_tag(self, infile):
train_sents = list(islice(brown.tagged(), 1000000))
trigram_tagger = tag.Trigram()
trigram_tagger.train(train_sents)
file = open(infile + ".txt", "r")
out = open(infile + "-tag.txt", "w")
try:
text = file.read()
lines = string.split(text, '\n')
for line in lines:
tokens = list(tokenize.whitespace(line))
tagged = list(trigram_tagger.tag(tokens))
for tags in tagged:
print tags
if tags[1] == None:
out.write(tags[0] + "/" + "NA")
else:
out.write(tags[0] + "/" + tags[1])
out.write(" ")
out.write("\n")
except IOError:
raise IOError()
file.close()
out.close()
from nltk import tag
from nltk.corpus import brown
import yaml
t0 = tag.Default("nn")
t1 = tag.Unigram(backoff=t0)
t1.train(brown.tagged("f")) # section a: press-reportage
f = open("demo_tagger.yaml", "w")
yaml.dump(t1, f)
from nltk import tag
from nltk.corpus import brown
import yaml
t0 = tag.Default('nn')
t1 = tag.Unigram(backoff=t0)
t1.train(brown.tagged('f')) # section a: press-reportage
f = open('demo_tagger.yaml', 'w')
yaml.dump(t1, f)
