# semantic parsing
if config.get('semantics', 'module') == 'boxer':
# tokenization
if not options.tokenized:
log.info("Tokenization with t")
tokens = candc.tokenize(text)
if not tokens:
log.error("error during tokenization of file '{0}', exiting".format(filename))
continue
tokenized = "\n".join([' '.join(sentence) for sentence in tokens[:-1]])
else:
tokenized = text
log.info("Parsing with Boxer")
semantics = candc.get_all(tokenized)
if not semantics:
log.error("error during the execution of Boxer on file '{0}', exiting".format(filename))
continue
elif config.get('semantics', 'module') == 'semafor':
log.info("Parsing with Semafor")
semantics, tokenized = semafor.parse(text)
if not semantics:
log.error("error during the execution of Semafor on file '{0}', exiting".format(filename))
continue
log.info("Word sense disambiguation and entity linking")
synsets, entities = disambiguation(tokenized, semantics)
if synsets==None or entities==None:
# tokenization
if not options.tokenized:
log.info("Tokenization")
tokens = tokenize(text)
if not tokens:
log.error(
"error during tokenization of file '{0}', exiting".format(
filename))
continue
tokenized = "\n".join([' '.join(sentence) for sentence in tokens[:-1]])
else:
tokenized = text
log.info("Parsing")
drs = get_all(tokenized)
if not drs:
log.error("error during the execution of Boxer on file '{0}', exiting".
format(filename))
continue
log.info("Word sense disambiguation and entity linking")
synsets, entities = disambiguation(tokenized, drs)
if synsets == None or entities == None:
log.error(
"error during the disambiguation of file '{0}', exiting".format(
filename))
continue
# extracting co-mentions
if options.comentions:
dbpedia_entities = set(map(lambda x: x['entity'], entities))
text = f.read()
# tokenization
if not options.tokenized:
log.info("Tokenization")
tokens = tokenize(text)
if not tokens:
log.error("error during tokenization of file '{0}', exiting".format(filename))
continue
tokenized = "\n".join([' '.join(sentence) for sentence in tokens[:-1]])
else:
tokenized = text
log.info("Parsing")
drs = get_all(tokenized)
if not drs:
log.error("error during the execution of Boxer on file '{0}', exiting".format(filename))
continue
log.info("Word sense disambiguation and entity linking")
synsets, entities = disambiguation(tokenized, drs)
if synsets==None or entities==None:
log.error("error during the disambiguation of file '{0}', exiting".format(filename))
continue
# extracting co-mentions
if options.comentions:
dbpedia_entities = set(map(lambda x: x['entity'], entities))
for entity1, entity2 in combinations(dbpedia_entities, 2):
if (entity1 != 'null' and
entity2 != 'null'):