def __init__(self, answer, i):
Feature.set_type(self, clas)
Feature.set_name(self, 'Sentence sentiment')
sentence = answer.sources[i].sentence
s = sum(map(lambda word: afinn.get(word, 0), [word.lower() for word in tokenize(sentence)]))
s = float(s) / len(sentence.split())
Feature.set_value(self, s)
def __init__(self, answer, i):
Feature.set_type(self, clas)
Feature.set_name(self, 'Question sentiment')
q = sum(
map(lambda word: afinn.get(word, 0),
[word.lower() for word in tokenize(answer.q.text)]))
q = float(q) / len(answer.q.text.split())
Feature.set_value(self, q)
def __init__(self, answer, i):
Feature.set_type(self, clas)
Feature.set_name(self, 'Sentence sentiment')
sentence = answer.sources[i].sentence
s = sum(
map(lambda word: afinn.get(word, 0),
[word.lower() for word in tokenize(sentence)]))
s = float(s) / len(sentence.split())
Feature.set_value(self, s)
def predict(self, answer):
s0 = ' '.join(tokenize(answer.q.text))
s1 = [' '.join(tokenize(source.sentence)) for source in answer.sources]
f = [source.features for source in answer.sources]
r = {
's0': s0,
's1': [dict([('text', s1_)] + [
(feat.get_type() + feat.get_name(), feat.get_value())
for feat in f_
])
for s1_, f_ in zip(s1, f)]
}
print(r)
req = urllib2.Request(self.url + 'score')
req.add_header('Content-Type', 'application/json')
resp = urllib2.urlopen(req, json.dumps(r))
rr = json.loads(resp.read())
print(rr)
return {'y': rr['score'], 'class': rr['class'], 'rel': rr['rel']}
def predict(self, answer):
s0 = ' '.join(tokenize(answer.q.text))
s1 = [' '.join(tokenize(source.sentence)) for source in answer.sources]
f = [source.features for source in answer.sources]
r = {
's0':
s0,
's1': [
dict([('text', s1_)] +
[(feat.get_type() + feat.get_name(), feat.get_value())
for feat in f_]) for s1_, f_ in zip(s1, f)
]
}
print(r)
req = urllib2.Request(self.url + 'score')
req.add_header('Content-Type', 'application/json')
resp = urllib2.urlopen(req, json.dumps(r))
rr = json.loads(resp.read())
print(rr)
return {'y': rr['score'], 'class': rr['class'], 'rel': rr['rel']}
def __init__(self, answer, i):
Feature.set_type(self, clas)
Feature.set_name(self, 'Question sentiment')
q = sum(map(lambda word: afinn.get(word, 0), [word.lower() for word in tokenize(answer.q.text)]))
q = float(q) / len(answer.q.text.split())
Feature.set_value(self, q)
