def create_candid_duration2date(doc):
"""
if a TimeX3 in the sentence is of type "duration", try to link it to a nearby type "date" TimeX3
if there is already another "date" in the same sentence, then skip this step
"""
candid = []
for s in doc.sents:
date = [tx for tx in s.timex3s if tx.type == 'DATE']
duration = [tx for tx in s.timex3s if tx.type == 'DURATION']
if not len(date) == 0 and len(duration) > 0:
print len(duration), 'duraion in sentence'
for i in range(1, 8):
pre_sent_num = s.num - i
if pre_sent_num >= 0:
d = [tx for tx in doc.sents[pre_sent_num].timex3s if tx.type == 'DATE']
if d != []:
for tx in duration:
tl = create_tlink_between_enty('dummy_id', d[0], tx, src='candid')
candid.append(tl)
break
post_sent_num = s.num + i
if post_sent_num < len(doc.sents):
d = [tx for tx in doc.sents[post_sent_num].timex3s if tx.type == 'DATE']
if d != []:
for tx in duration:
tl = create_tlink_between_enty('dummy_id', tx, d[0], src='candid')
candid.append(tl)
break
return candid
def expand(s):
inferable_pair = [tpl[0:2] for tpl in accepted_state]
uncertain_pair = [tpl[0:2] for tpl in unknown_state]
infered = []
sorted_entits = sorted(s.events + s.timex3s, key=lambda x: x.span[0].begin)
l = len(sorted_entits)
for r in range(2, l):
for fi in range(0, l - r):
from_enty = sorted_entits[fi]
ti = fi + r
to_enty = sorted_entits[ti]
tl3 = create_tlink_between_enty('dummy_id', from_enty, to_enty, src='expand')
tl3.votes = {'BEFORE': 0, 'OVERLAP': 0, 'AFTER': 0}
for d in range(1, r):
mi = fi + d
mid_enty = sorted_entits[mi]
tl1 = search_tlink_between_enty(from_enty, mid_enty, s.candids_within)
tl2 = search_tlink_between_enty(mid_enty, to_enty, s.candids_within)
if tl1 and tl2:
if (tl1.pred, tl2.pred) in inferable_pair[0:3]:
tl3.votes['BEFORE'] += 1
if (tl1.pred, tl2.pred) in inferable_pair[3:6]:
tl3.votes['AFTER'] += 1
if (tl1.pred, tl2.pred) in uncertain_pair:
tl3.votes['OVERLAP'] += 1
else:
print 'some bridge tlink is not ready'
if sum(tl3.votes.values()) > 0:
tl3.pred = get_kth_large_key(tl3.votes)
infered.append(tl3)
s.candids_within.append(tl3)
def create_candid_within3(doc):
candids = []
for s in doc.sents:
s.candids_within = []
sorted_entits = sorted(s.events + s.timex3s, key=lambda x: x.span[0].begin)
'''
s.freq_tl = []
s.freq_tx = []
for tx in sorted(s.timex3s, key=lambda x: x.span[0].begin):
if tx.type == 'FREQUENCY':
s.freq_tx.append(tx)
for tx in s.freq_tx:
i = sorted_entits.index(tx)
for d in range(1, len(sorted_entits)):
if i - d >= 0:
e = sorted_entits[i - d]
if e.sig == 'Event':
tl = create_tlink_between_enty('dummy_id', e, tx, src='candid')
s.freq_tl.append(tl)
#s.candids_within.append(tl)
#candids.append(tl)
sorted_entits.remove(tx)
'''
# generate within sentence candidate
for pair in combinations(sorted_entits, 2):
tl = create_tlink_between_enty('dummy_id', pair[0], pair[1], src='candid')
s.candids_within.append(tl)
candids.append(tl)
return candids
def create_candid_within2(doc):
candids = []
for s in doc.sents:
s.candids_within = []
# generate within sentence candidate
sorted_entits = sorted(s.events + s.timex3s, key=lambda x: x.span[0].begin)
for i in range(len(sorted_entits) - 1):
tl = create_tlink_between_enty('dummy_id', sorted_entits[i], sorted_entits[i + 1], src='candid')
s.candids_within.append(tl)
candids.append(tl)
#for i in range(len(sorted_entits) - 1):
# tl = create_tlink_between_enty('dummy_id', sorted_entits[i], sorted_entits[i + 1], src='candid')
# candid.append(tl)
return candids
def create_candid_coref(doc):
sent_src = doc.sents[:]
candid = []
for i in range(len(sent_src) - 1):
fs = sent_src[i]
fs_entits = fs.events
for ts in sent_src[i + 1:]:
ts_entits = ts.events
for pair in product(fs_entits, ts_entits):
if get_phrase_head(pair[0]).lemma == get_phrase_head(pair[1]).lemma:
tl = create_tlink_between_enty('dummy_id', pair[0], pair[1])
tl.pred = 'OVERLAP'
candid.append(tl)
return candid
def enumerate_possible_between(doc):
sent_src = doc.sents[:]
for i in [doc.admission_event, doc.admission_timex3,
doc.discharge_event, doc.discharge_timex3]:
if i:
sent_src.remove(i.sent)
ptls = []
ptl_counter = 0
for i in range(len(sent_src) - 1):
fs = sent_src[i]
fs_entits = fs.events + fs.timex3s
for ts in sent_src[i + 1:]:
ts_entits = ts.events + ts.timex3s
for pair in product(fs_entits, ts_entits):
tl = create_tlink_between_enty(doc.ds_id + '_ptl_' + str(ptl_counter), pair[0], pair[1])
ptls.append(tl)
ptl_counter += 1
return ptls
def create_candid_event2sectime(doc):
event_src = doc.events[:]
if doc.admission_event:
event_src.remove(doc.admission_event)
if doc.discharge_event:
event_src.remove(doc.discharge_event)
candid = []
for e in event_src:
if e.sent.sect == 0:
if doc.admission_timex3:
tl = create_tlink_between_enty('dummy_id', doc.admission_timex3, e, src='candid')
else:
tl = create_tlink_between_enty('dummy_id', doc.admission_event, e, src='candid')
if e.text.lower() == 'admission':
tl.pred = 'OVERLAP'
else:
tl.pred = 'AFTER'
candid.append(tl)
if e.sent.sect == 1:
if doc.discharge_timex3:
tl = create_tlink_between_enty('dummy_id', doc.discharge_timex3, e, src='candid')
else:
tl = create_tlink_between_enty('dummy_id', doc.discharge_event, e, src='candid')
if 'discharge' in e.text.lower():
tl.pred = 'OVERLAP'
else:
tl.pred = 'AFTER'
candid.append(tl)
if doc.admission_event and doc.admission_timex3:
tl = create_tlink_between_enty('dummy_id', doc.admission_event, doc.admission_timex3, src='candid')
tl.pred = 'OVERLAP'
candid.append(tl)
if doc.discharge_event and doc.discharge_timex3:
tl = create_tlink_between_enty('dummy_id', doc.discharge_event, doc.discharge_timex3, src='candid')
tl.pred = 'OVERLAP'
candid.append(tl)
return candid
