import Parser, scikit_classifier
if __name__ == '__main__':
# takes forever to run
with open('data/output/searchresult.txt', 'w') as f:
for ngram_size in [0, 1, 2]:
for use_syntactic_features in [0]:
for pos_window_size in [0, 1, 5, 10]:
for window_size in [0, 5, 10, 50, 100, 500]:
for use_lesk in [False]:
for use_lesk_words in [False]:
classifier = scikit_classifier.scikit_classifier(
pos_window_size=pos_window_size,
ngram_size=ngram_size,
window_size=window_size,
use_syntactic_features=
use_syntactic_features,
use_lesk=use_lesk,
use_lesk_words=use_lesk_words)
egs = Parser.load_examples(
'data/wsd-data/train_split.data')
test_egs = Parser.load_examples(
'data/wsd-data/valiation_split.data')
classifier.train(egs)
tp = 0.0
fp = 0.0
tn = 0.0
fn = 0.0
for eg in test_egs:
import Parser, scikit_classifier
if __name__ == '__main__':
# takes forever to run
with open('data/output/searchresult.txt', 'w') as f:
for ngram_size in [0,1,2]:
for use_syntactic_features in [0]:
for pos_window_size in [0,1,5,10]:
for window_size in [0, 5, 10, 50, 100, 500]:
for use_lesk in [False]:
for use_lesk_words in [False]:
classifier = scikit_classifier.scikit_classifier(
pos_window_size = pos_window_size,
ngram_size = ngram_size,
window_size = window_size,
use_syntactic_features = use_syntactic_features,
use_lesk = use_lesk,
use_lesk_words = use_lesk_words)
egs = Parser.load_examples('data/wsd-data/train_split.data')
test_egs = Parser.load_examples('data/wsd-data/valiation_split.data')
classifier.train(egs)
tp = 0.0
fp = 0.0
tn = 0.0
fn = 0.0
for eg in test_egs:
pred = classifier.predict([eg])
for (k,(s,p)) in enumerate(zip(eg.senses,pred)):
if s == 1 and p == 1:
#testdata = Parser.load_data('data/wsd-data/test.data')
testdata = Parser.load_test_data('data/wsd-data/test.data')
with open('data/output/%s.txt'%name, 'w') as f:
for testexample in testdata:
prediction = classifier.predict([testexample])
for element in prediction:
f.write('%d\n' % element)
f.close()
print("done")
elif task == 5:
name = "scikit_test_ws500_pos1_ngram0_synfeat0_uselesk_useleskwords"
classifier = scikit_classifier.scikit_classifier(
window_size=500,
use_syntactic_features=0,
pos_window_size=1,
ngram_size=0,
use_lesk=True,
use_lesk_words=True,
training_file='data/wsd-data/train.data',
test_file='data/wsd-data/test.data')
egs = Parser.load_examples('data/wsd-data/train.data')
test_egs = Parser.load_examples('data/wsd-data/test.data')
# Train the classifier(s)
classifier.train(egs)
with open('data/output/%s.txt'%name, 'w') as f:
for eg in test_egs:
pred = classifier.predict([eg])
for p in pred:
f.write("%d\n"%p)
#testdata = Parser.load_data('data/wsd-data/test.data')
testdata = Parser.load_test_data('data/wsd-data/test.data')
with open('data/output/%s.txt' % name, 'w') as f:
for testexample in testdata:
prediction = classifier.predict([testexample])
for element in prediction:
f.write('%d\n' % element)
f.close()
print("done")
elif task == 5:
name = "scikit_test_ws500_pos1_ngram0_synfeat0_uselesk_useleskwords"
classifier = scikit_classifier.scikit_classifier(
window_size=500,
use_syntactic_features=0,
pos_window_size=1,
ngram_size=0,
use_lesk=True,
use_lesk_words=True,
training_file='data/wsd-data/train.data',
test_file='data/wsd-data/test.data')
egs = Parser.load_examples('data/wsd-data/train.data')
test_egs = Parser.load_examples('data/wsd-data/test.data')
# Train the classifier(s)
classifier.train(egs)
with open('data/output/%s.txt' % name, 'w') as f:
for eg in test_egs:
pred = classifier.predict([eg])
for p in pred:
f.write("%d\n" % p)