def do_extract(self):
freq_port = FreqPortion(self.freq_data, self.selected_doc)
phrase_freq = freq_port.compute_freq_portion()
normalize_feature(phrase_freq)
context_freq = ContextFreq(self.freq_data, self.selected_doc, self.selected_context, self.total_cnt)
phrase_context_freq = context_freq.compute_context_freq()
normalize_feature(phrase_context_freq)
context_freq = ContextScore(self.freq_data, self.selected_doc, self.selected_context, self.total_cnt, self.global_scores)
phrase_context_freq = context_freq.compute_context_score()
#normalize_feature(phrase_context_freq)
idf = IDFMeasure(self.freq_data, self.selected_doc)
phrase_idf = idf.compute_idf()
normalize_feature(phrase_idf)
cap = CapPortion(self.freq_data, self.parsed_file, self.selected_doc)
phrase_cap_portion = cap.compute_cap_portion()
normalize_feature(phrase_cap_portion)
punc = PuncPortion(self.freq_data, self.parsed_file, self.selected_doc)
phrase_punc_portion = punc.compute_punc_portion()
normalize_feature(phrase_punc_portion)
mi_kl = MIKL(self.freq_patterns, self.freq_data, self.selected_doc)
phrase_mi, phrase_kl = mi_kl.compute_mi_and_kl()
normalize_feature(phrase_mi)
normalize_feature(phrase_kl)
# Experimental
#context_freq_dist = ContextFreqDist(self.freq_data, self.selected_doc, self.selected_context, self.total_cnt)
#phrase_context_freq_dist = context_freq_dist.compute_context_freq_dist()
return [phrase_freq, phrase_idf, phrase_cap_portion, phrase_punc_portion, phrase_mi, phrase_kl, phrase_context_freq]
def do_extract(self):
freq_port = FreqPortion(self.freq_data, self.selected_doc)
phrase_freq = freq_port.compute_freq_portion()
normalize_feature(phrase_freq)
idf = IDFMeasure(self.freq_data, self.selected_doc)
phrase_idf = idf.compute_idf()
normalize_feature(phrase_idf)
cap = CapPortion(self.freq_data, self.parsed_file, self.selected_doc)
phrase_cap_portion = cap.compute_cap_portion()
normalize_feature(phrase_cap_portion)
punc = PuncPortion(self.freq_data, self.parsed_file, self.selected_doc)
phrase_punc_portion = punc.compute_punc_portion()
normalize_feature(phrase_punc_portion)
mi_kl = MIKL(self.freq_patterns, self.freq_data, self.selected_doc)
phrase_mi, phrase_kl = mi_kl.compute_mi_and_kl()
normalize_feature(phrase_mi)
normalize_feature(phrase_kl)
return [phrase_freq, phrase_idf, phrase_cap_portion, phrase_punc_portion, phrase_mi, phrase_kl]
def calc_punc_portion(doc, freq_data): punc_portion = PuncPortion(freq_data, doc) print punc_portion.compute_punc_portion()