def classify(context):
count = 0
try:
m = load_model()
for cont in context:
print_progress(count, len(context))
count += 1
try:
line_split = cont.split("\t")
id = line_split[0]
sent = line_split[1]
word = line_split[2]
if os.path.isfile("../test/subtask3/results_" + str(id) +
".txt"):
continue
original_sense, other_sense = produce_synset(sent, word, m)
# print original_sense, other_sense
original_sense_key = original_sense[0][1].lemmas()[0].key()
other_sense_key = other_sense[0][1].lemmas()[0].key()
with open("../test/subtask3/results_" + str(id) + ".txt",
'w') as out:
out.write(
str(id) + " " + original_sense_key + " " +
other_sense_key + " " + original_sense[0][2] + " " +
other_sense[0][2])
except Exception as e:
pass
except Exception as e:
print e
def subtask3_topkeys():
size, count = 10, 0
lines = open("../data/subtask3-test.tsv").read().split("\n")
size = len(lines)
# m = load_model()
for line in lines[2:3]:
count += 1
# try:
line_split = line.split("\t")
id = line_split[0]
print id
sent = line_split[1]
word = line_split[2]
# if os.path.isfile("../baselines/subtask3/results_" + str(id) + ".txt"):
# continue
try:
with open("../baselines/subtask3/results_" + str(id) + ".txt",
'w') as out:
original_sense, other_sense = synset_baseline(sent, word)
original_sense = wn.synsets(
original_sense)[0].lemmas()[0].key()
other_sense = wn.synsets(other_sense)[0]
other_sense = other_sense.lemmas()[0].key()
out.write(str(id) + " " + original_sense + " " + other_sense)
except Exception as e:
raise
print_progress(count, size)
def classify_linear():
size, count = 10, 0
lines = open("../data/subtask3-test.tsv").read().split("\n")
size = len(lines)
# m = load_model()
for line in lines:
count += 1
# try:
line_split = line.split("\t")
id = line_split[0]
sent = line_split[1]
word = line_split[2]
if os.path.isfile("../baselines/subtask3/results_" + str(id) + ".txt"):
continue
with open("../baselines/subtask3/results_" + str(id) + ".txt",
'w') as out:
original_sense, other_sense = synset_baseline(sent, word)
# print original_sense, other_sense
original_sense_key = original_sense[0][1].lemmas()[0].key()
other_sense_key = other_sense[0][1].lemmas()[0].key()
out.write(
str(id) + " " + original_sense_key + " " + other_sense_key +
" " + original_sense[0][2] + " " + other_sense[0][2])
# except Exception as e:
# print e
# pass
print_progress(count, size)
id = context[1]
if os.path.isfile("../baselines/subtask1/results_" + str(id) + ".txt"):
return
try:
res = "1" if is_pun(context[0]) else "0"
with open("../baselines/subtask1/results_" + str(id) + ".txt",
'w') as f:
f.write(id + " " + res)
return
except Exception as e:
return
if __name__ == "__main__":
# size = 10
puns = []
tsv_file = open("../data/hetero_annotations.tsv").read()
size = len(tsv_file.split("\n"))
for i, line in enumerate(tsv_file.split("\n")[1:size + 1]):
sent = line.split("\t")[0]
puns.append((sent, "het_" + str(i)))
count = 0
print_progress(count, size)
pool = multiprocessing.Pool(processes=4)
pun_it = pool.imap_unordered(classify_pun, puns)
for res in pun_it:
count += 1
print_progress(count, size)
import multiprocessing
from data_processing import print_progress
from pun_word_4grams import pun_data
def write_data(line):
line_split = line.split("\t")
# print line_split
if len(line_split) != 6:
return
try:
with open("../data/subtask3_pun_data/" + line_split[0] + ".json", 'w') as out:
json.dump(pun_data(line_split[1], line_split[2]), out, indent=4)
except Exception as e:
print e
print "Failed on", line_split[1]
if __name__ == "__main__":
with open("../data/subtask3-annotations-full-2.tsv") as f:
data = f.read().split("\n")
count = 0
pool = multiprocessing.Pool(processes=4)
pun_it = pool.imap_unordered(write_data, data)
print_progress(count, len(data))
for res in pun_it:
count += 1
print_progress(count, len(data))
