def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = search_context.topic.get_topic_text()
if self.topic_lang_model is None:
self.topic_lang_model = self._generate_topic_language_model(
search_context)
snip_text = self._get_snip_text(search_context)
all_text = topic_text + ' ' + snip_text
all_text = self._check_terms(all_text)
bi_query_generator = BiTermQueryGeneration(
minlen=3, stopwordfile=self._stopword_file)
tri_query_generator = TriTermQueryGeneration(
minlen=3, stopwordfile=self._stopword_file)
tri_query_list = tri_query_generator.extract_queries_from_text(
all_text)
bi_query_list = bi_query_generator.extract_queries_from_text(all_text)
query_list = tri_query_list + bi_query_list
query_ranker = QueryRanker(
smoothed_language_model=self.topic_lang_model)
query_ranker.calculate_query_list_probabilities(query_list)
gen_query_list = query_ranker.get_top_queries(100)
return gen_query_list
def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = search_context.topic.get_topic_text()
if self.topic_lang_model is None:
self.topic_lang_model = self._generate_topic_language_model(search_context)
snip_text = self._get_snip_text(search_context)
all_text = topic_text + ' ' + snip_text
all_text = self._check_terms(all_text)
bi_query_generator = BiTermQueryGeneration(minlen=3, stopwordfile=self._stopword_file)
tri_query_generator = TriTermQueryGeneration(minlen=3, stopwordfile=self._stopword_file)
tri_query_list = tri_query_generator.extract_queries_from_text(all_text)
bi_query_list = bi_query_generator.extract_queries_from_text(all_text)
query_list = tri_query_list + bi_query_list
query_ranker = QueryRanker(smoothed_language_model=self.topic_lang_model)
query_ranker.calculate_query_list_probabilities(query_list)
gen_query_list = query_ranker.get_top_queries(100)
return gen_query_list
def _rank_terms(self, terms, **kwargs):
"""
Ranks the query terms by their discriminatory power.
The length of the list returned == list of initial terms supplied.
"""
topic_language_model = kwargs.get('topic_language_model', None)
ranker = QueryRanker(smoothed_language_model=topic_language_model)
ranker.calculate_query_list_probabilities(terms)
return ranker.get_top_queries(len(terms))
def _rank_terms(self, terms, **kwargs):
"""
Ranks the query terms by their discriminatory power.
The length of the list returned == list of initial terms supplied.
"""
topic_language_model = kwargs.get('topic_language_model', None)
ranker = QueryRanker(smoothed_language_model=topic_language_model)
ranker.calculate_query_list_probabilities(terms)
return ranker.get_top_queries(len(terms))
def generate_query_list(self, topic):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = topic.content
topic_lang_model = self._generate_topic_language_model(topic)
single_query_generator = SingleQueryGeneration(minlen=3, stopwordfile=self._stopword_file)
single_query_list = single_query_generator.extract_queries_from_text(topic_text)
query_ranker = QueryRanker(smoothed_language_model=topic_lang_model)
query_ranker.calculate_query_list_probabilities(single_query_list)
return query_ranker.get_top_queries(100)
def produce_query_list(self, topic):
topic_text = topic.content
topicLM = self.make_topic_lm(topic)
bi_query_generator = BiTermQueryGeneration(minlen=3, stopwordfile=self.stopword_file)
tri_query_generator = TriTermQueryGeneration(minlen=3, stopwordfile=self.stopword_file)
tri_query_list = tri_query_generator.extract_queries_from_text(topic_text)
bi_query_list = bi_query_generator.extract_queries_from_text(topic_text)
query_list = tri_query_list + bi_query_list
qr = QueryRanker(smoothed_language_model=topicLM)
qr.calculate_query_list_probabilities(query_list)
queries = qr.get_top_queries(100)
return queries
def generate_query_list(self, topic):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = topic.content
topic_lang_model = self._generate_topic_language_model(topic)
single_query_generator = SingleQueryGeneration(
minlen=3, stopwordfile=self._stopword_file)
single_query_list = single_query_generator.extract_queries_from_text(
topic_text)
query_ranker = QueryRanker(smoothed_language_model=topic_lang_model)
query_ranker.calculate_query_list_probabilities(single_query_list)
return query_ranker.get_top_queries(100)
def produce_query_list(self, topic):
topic_text = topic.content
topicLM = self.make_topic_lm(topic)
bi_query_generator = BiTermQueryGeneration(
minlen=3, stopwordfile=self.stopword_file)
tri_query_generator = TriTermQueryGeneration(
minlen=3, stopwordfile=self.stopword_file)
tri_query_list = tri_query_generator.extract_queries_from_text(
topic_text)
bi_query_list = bi_query_generator.extract_queries_from_text(
topic_text)
query_list = tri_query_list + bi_query_list
qr = QueryRanker(smoothed_language_model=topicLM)
qr.calculate_query_list_probabilities(query_list)
queries = qr.get_top_queries(100)
return queries
def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic = search_context.topic
topic_text = "{0} {1}".format(topic.title, topic.content)
topic_language_model = self._generate_topic_language_model(search_context)
generator = SingleQueryGeneration(minlen=3, stopwordfile=self._stopword_file)
query_list = generator.extract_queries_from_text(topic_text)
query_ranker = QueryRanker(smoothed_language_model=topic_language_model)
query_ranker.calculate_query_list_probabilities(query_list)
generated_queries = query_ranker.get_top_queries(100)
return generated_queries
def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = search_context.topic.get_topic_text()
if self.topic_lang_model is None:
self.topic_lang_model = self._generate_topic_language_model(search_context)
snip_text = self._get_snip_text(search_context)
all_text = topic_text + ' ' + snip_text
all_text = self._check_terms(all_text)
term_list = all_text.split(' ')
term_list = list(set(term_list))
q3_list = list(itertools.combinations(term_list,3))
q4_list = list(itertools.combinations(term_list,4))
query_list = []
for q in q3_list:
query_list.append( ' '.join(q))
for q in q4_list:
query_list.append( ' '.join(q))
query_ranker = QueryRanker(smoothed_language_model=self.topic_lang_model)
query_ranker.calculate_query_list_probabilities(query_list)
gen_query_list = query_ranker.get_top_queries(100)
return gen_query_list
def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic = search_context.topic
topic_text = "{0} {1}".format(topic.title, topic.content)
topic_lang_model = self._generate_topic_language_model(search_context)
bi_query_generator = BiTermQueryGeneration(
minlen=3, stopwordfile=self._stopword_file)
bi_query_list = bi_query_generator.extract_queries_from_text(
topic_text)
query_list = bi_query_list
query_ranker = QueryRanker(smoothed_language_model=topic_lang_model)
query_ranker.calculate_query_list_probabilities(query_list)
gen_query_list = query_ranker.get_top_queries(100)
return gen_query_list
def generate_query_list(self, search_context):
"""
Given a Topic object, produces a list of query terms that could be issued by the simulated agent.
"""
topic_text = search_context.topic.get_topic_text()
if self.topic_lang_model is None:
self.topic_lang_model = self._generate_topic_language_model(
search_context)
snip_text = self._get_snip_text(search_context)
all_text = topic_text + ' ' + snip_text
all_text = self._check_terms(all_text)
term_list = all_text.split(' ')
term_list = list(set(term_list))
q3_list = list(itertools.combinations(term_list, 3))
q4_list = list(itertools.combinations(term_list, 4))
query_list = []
for q in q3_list:
query_list.append(' '.join(q))
for q in q4_list:
query_list.append(' '.join(q))
query_ranker = QueryRanker(
smoothed_language_model=self.topic_lang_model)
query_ranker.calculate_query_list_probabilities(query_list)
gen_query_list = query_ranker.get_top_queries(100)
return gen_query_list
