def print_sparql_queries():
import argparse
parser = argparse.ArgumentParser(description="Dump qa entity pairs.")
parser.add_argument("--config",
default="config.cfg",
help="The configuration file to use.")
parser.add_argument("--output",
help="The file to dump results to.")
args = parser.parse_args()
globals.read_configuration(args.config)
scorer_globals.init()
parameters = translator.TranslatorParameters()
parameters.require_relation_match = False
parameters.restrict_answer_type = False
dataset = "webquestions_test_filter"
sparql_backend = globals.get_sparql_backend(globals.config)
queries = get_evaluated_queries(dataset, True, parameters)
for index, query in enumerate(queries):
print "--------------------------------------------"
print query.utterance
print "\n".join([str((entity.__class__, entity.entity)) for entity in query.eval_candidates[0].query_candidate.query.identified_entities])
for eval_candidate in query.eval_candidates:
query_candidate = eval_candidate.query_candidate
query_candidate.sparql_backend = sparql_backend
notable_types = query_candidate.get_answers_notable_types()
if notable_types:
print notable_types
print query_candidate.graph_as_simple_string().encode("utf-8")
print query_candidate.to_sparql_query().encode("utf-8")
print "\n\n"
def init_from_config():
config_params = globals.config
sparql_backend = globals.get_sparql_backend(config_params)
query_extender = QueryCandidateExtender.init_from_config()
entity_linker = EntityLinker.init_from_config()
parser = CoreNLPParser.init_from_config()
scorer_obj = ranker.SimpleScoreRanker('DefaultScorer')
return QueryTranslator(sparql_backend, query_extender, entity_linker,
parser, scorer_obj)
def init_from_config():
config_params = globals.config
sparql_backend = globals.get_sparql_backend(config_params)
query_extender = QueryCandidateExtender.init_from_config()
entity_linker = EntityLinker.init_from_config()
parser = CoreNLPParser.init_from_config()
scorer_obj = ranker.SimpleScoreRanker('DefaultScorer')
return QueryTranslator(sparql_backend, query_extender,
entity_linker, parser, scorer_obj)
def init_from_config(args):
global w2v, sparql_backend, entity_linker, facts_ranker, facts_extractor
global wiki_url
config_options = globals.config
w2v = Word2Vec.init_from_config(config_options)
sparql_backend = globals.get_sparql_backend(config_options)
wiki_url = WikiUrl(config_options)
entity_linker = EntityLinker.init_from_config(config_options, wiki_url)
facts_ranker = Ranker.init_from_config(config_options)
facts_extractor = FactExtractor.init_from_config(config_options)
def get_from_config(cls, config_params):
sparql_backend = globals.get_sparql_backend(config_params)
query_extender = QueryCandidateExtender.init_from_config()
entity_linker = globals.get_entity_linker()
parser = globals.get_parser()
scorer_obj = ranker.SimpleScoreRanker('DefaultScorer')
ngram_notable_types_npmi_path = config_params.get('QueryCandidateExtender', 'ngram-notable-types-npmi', '')
notable_types_npmi_threshold = float(config_params.get('QueryCandidateExtender', 'notable-types-npmi-threshold'))
ngram_notable_types_npmi = None
if ngram_notable_types_npmi_path and os.path.exists(ngram_notable_types_npmi_path):
import cPickle as pickle
try:
with open(ngram_notable_types_npmi_path, 'rb') as inp:
logger.info("Loading types model from disk...")
ngram_notable_types_npmi = pickle.load(inp)
except IOError as exc:
logger.error("Error reading types model: %s" % str(exc))
ngram_notable_types_npmi = None
return SparqlQueryTranslator(sparql_backend, query_extender,
entity_linker, parser, scorer_obj,
ngram_notable_types_npmi,
notable_types_npmi_threshold)
import globals
import sys
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(description="Dump qa entity pairs.")
parser.add_argument("--config",
default="config.cfg",
help="The configuration file to use.")
args = parser.parse_args()
globals.read_configuration(args.config)
sparql_backend = globals.get_sparql_backend(globals.config)
while True:
print "Please enter query: "
query_str = ""
while True:
query_str_line = sys.stdin.readline()
if query_str_line.startswith("END"):
break
query_str += query_str_line
print sparql_backend.query(query_str)
def init_from_config():
"""
Return an instance with options parsed by a config parser.
:param config_options:
:return:
"""
config_options = globals.config
sparql_backend = globals.get_sparql_backend(config_options)
relation_counts_file = config_options.get('QueryCandidateExtender',
'relation-counts')
mediator_names_file = config_options.get('QueryCandidateExtender',
'mediator-names')
reverse_relations_file = config_options.get('QueryCandidateExtender',
'reverse-relations')
expected_types_file = config_options.get('QueryCandidateExtender',
'relation-expected-types')
tt_distributions_file = config_options.get('QueryCandidateExtender',
'relation-target-type-distributions')
mediator_relations_file = config_options.get('QueryCandidateExtender',
'mediator-relations')
rel_lemmas_file = config_options.get('QueryCandidateExtender',
'relation-lemmas')
relation_words_file = config_options.get('QueryCandidateExtender',
'relation-words')
mediated_relation_words_file = config_options.get(
'QueryCandidateExtender',
'mediated-relation-words')
word_type_counts_file = config_options.get(
'QueryCandidateExtender',
'word-type-counts')
word_type_counts = data.read_word_type_distributions(
word_type_counts_file)
embeddings_model = config_options.get('Alignment',
'word-embeddings')
word_deriv_file = config_options.get('Alignment',
'word-derivations')
we_synonyms = WordembeddingSynonyms(embeddings_model)
word_derivations = WordDerivations(word_deriv_file)
mediator_relations = data.read_mediator_relations(
mediator_relations_file)
relation_counts = data.read_relation_counts(relation_counts_file)
mediator_names = data.read_mediator_names(mediator_names_file)
mediator_index = MediatorIndexFast.init_from_config()
reverse_relations = data.read_reverse_relations(reverse_relations_file)
relation_expected_types = data.read_relation_expected_types(
expected_types_file)
relation_words = data.read_relation_words(relation_words_file,
n_top_words=1000)
mediated_relation_words = data.read_mediated_relation_words(
mediated_relation_words_file, n_top_words=1000)
rel_tt_distributions = data.read_relation_target_type_distributions(
tt_distributions_file)
rel_lemmas = data.read_relation_lemmas(rel_lemmas_file)
return QueryCandidateExtender(mediator_index, relation_counts,
mediator_names,
mediator_relations,
reverse_relations,
relation_expected_types,
sparql_backend, relation_words,
mediated_relation_words,
rel_tt_distributions, we_synonyms,
word_derivations, word_type_counts,
rel_lemmas)