def fillEntry_ref(self, arnumber, layer):
if not arnumber in self.data:
self.data[arnumber] = {}
self.data[arnumber]['refs'] = getRefs(arnumber)
self.data[arnumber]['keywords'] = index(arnumber)
#index function return "IEEE TERMS" of paper ID arnumber
self.data[arnumber]['layer'] = layer
def getNextLayer_cite(self, curr_layer_lst):
tmp = {}
for num in curr_layer_lst:
for cite in self.data[num]["cites"]:
if not cite in self.data:
if not cite in tmp:
citation = len(getCites(cite))
if citation > 10:
tmp[cite] = {}
tmp[cite]["citation"] = citation
tmp[cite]["keywords"] = index(cite)
tmp[cite]["similarity"] = len(set(self.data[num]["keywords"]) & set(tmp[cite]["keywords"]))
tmp[cite]["repeat"] = 1
else:
siml = len(set(self.data[num]["keywords"]) & set(tmp[cite]["keywords"]))
tmp[cite]["similarity"] = max(tmp[cite]["similarity"], siml)
tmp[cite]["repeat"] += 1
score = {}
for cite in tmp:
score[cite] = tmp[cite]["similarity"] * tmp[cite]["citation"] * tmp[cite]["repeat"]
score = sorted(score.iteritems(), key=operator.itemgetter(1), reverse=True)
maxIteration = min(len(score), self.nodes_cite)
result = [0] * maxIteration
for i in xrange(maxIteration):
result[i] = score[i][0]
print result[i]
print tmp[result[i]]["similarity"], tmp[result[i]]["citation"], tmp[result[i]]["repeat"]
return result
def getNextLayer_ref(self, curr_layer_lst):
tmp = {}
for num in curr_layer_lst:
print "%d is processed" % (num)
for ref in self.data[num]["refs"]:
print " refer paper %d is accessed" % (ref)
if not ref in self.data:
if not ref in tmp:
citation = len(getCites(ref))
if citation > 1:
tmp[ref] = {}
tmp[ref]["citation"] = citation
tmp[ref]["keywords"] = index(ref)
tmp[ref]["similarity"] = len(set(self.data[num]["keywords"]) & set(tmp[ref]["keywords"]))
tmp[ref]["repeat"] = 1
else:
siml = len(set(self.data[num]["keywords"]) & set(tmp[ref]["keywords"]))
tmp[ref]["similarity"] = max(tmp[ref]["similarity"], siml)
tmp[ref]["repeat"] += 1
score = {}
training_data = []
print len(tmp)
for ref in tmp:
score[ref] = tmp[ref]["similarity"] * tmp[ref]["citation"] * tmp[ref]["repeat"]
# Candidate ref papers are given scores by similarity * citation * # of citations from last layer
training_data.append([tmp[ref]["similarity"], tmp[ref]["citation"], tmp[ref]["repeat"]])
score = sorted(score.iteritems(), key=operator.itemgetter(1), reverse=True)
maxIteration = min(len(score), self.nodes_ref)
result = [0] * maxIteration
for i in xrange(maxIteration):
result[i] = score[i][0]
print result[i]
print tmp[result[i]]["similarity"], tmp[result[i]]["citation"], tmp[result[i]]["repeat"]
# Top nodes_ref papers are chosen for next layer
return result, training_data
def fillEntry_ref(self, arnumber, layer):
if not arnumber in self.data:
self.data[arnumber] = {}
self.data[arnumber]["refs"] = getRefs(arnumber)
self.data[arnumber]["keywords"] = index(arnumber)
# index function return "IEEE TERMS" of paper ID arnumber
self.data[arnumber]["layer"] = layer
def getNextLayer_ref(self, curr_layer_lst):
tmp = {}
for num in curr_layer_lst:
print "%d is processed" % (num)
for ref in self.data[num]['refs']:
print " refer paper %d is accessed" % (ref)
if not ref in self.data:
if not ref in tmp:
citation = len(getCites(ref))
if citation > 1:
tmp[ref] = {}
tmp[ref]['citation'] = citation
tmp[ref]['keywords'] = index(ref)
tmp[ref]['similarity'] = len(
set(self.data[num]['keywords'])
& set(tmp[ref]['keywords']))
tmp[ref]['repeat'] = 1
else:
siml = len(
set(self.data[num]['keywords'])
& set(tmp[ref]['keywords']))
tmp[ref]['similarity'] = max(tmp[ref]['similarity'],
siml)
tmp[ref]['repeat'] += 1
score = {}
training_data = []
print len(tmp)
for ref in tmp:
score[ref] = tmp[ref]['similarity'] * tmp[ref]['citation'] * tmp[
ref]['repeat']
# Candidate ref papers are given scores by similarity * citation * # of citations from last layer
training_data.append([
tmp[ref]['similarity'], tmp[ref]['citation'],
tmp[ref]['repeat']
])
score = sorted(score.iteritems(),
key=operator.itemgetter(1),
reverse=True)
maxIteration = min(len(score), self.nodes_ref)
result = [0] * maxIteration
for i in xrange(maxIteration):
result[i] = score[i][0]
print result[i]
print tmp[result[i]]['similarity'], tmp[
result[i]]['citation'], tmp[result[i]]['repeat']
# Top nodes_ref papers are chosen for next layer
return result, training_data
def getNextLayer_cite(self, curr_layer_lst):
tmp = {}
for num in curr_layer_lst:
for cite in self.data[num]['cites']:
if not cite in self.data:
if not cite in tmp:
citation = len(getCites(cite))
if citation > 10:
tmp[cite] = {}
tmp[cite]['citation'] = citation
tmp[cite]['keywords'] = index(cite)
tmp[cite]['similarity'] = len(
set(self.data[num]['keywords'])
& set(tmp[cite]['keywords']))
tmp[cite]['repeat'] = 1
else:
siml = len(
set(self.data[num]['keywords'])
& set(tmp[cite]['keywords']))
tmp[cite]['similarity'] = max(tmp[cite]['similarity'],
siml)
tmp[cite]['repeat'] += 1
score = {}
for cite in tmp:
score[cite] = tmp[cite]['similarity'] * tmp[cite][
'citation'] * tmp[cite]['repeat']
score = sorted(score.iteritems(),
key=operator.itemgetter(1),
reverse=True)
maxIteration = min(len(score), self.nodes_cite)
result = [0] * maxIteration
for i in xrange(maxIteration):
result[i] = score[i][0]
print result[i]
print tmp[result[i]]['similarity'], tmp[
result[i]]['citation'], tmp[result[i]]['repeat']
return result
def fillEntry_cite(self, arnumber, layer):
if not arnumber in self.data:
self.data[arnumber] = {}
self.data[arnumber]['cites'] = getCites(arnumber)
self.data[arnumber]['keywords'] = index(arnumber)
self.data[arnumber]['layer'] = layer
def fillEntry_cite(self, arnumber, layer):
if not arnumber in self.data:
self.data[arnumber] = {}
self.data[arnumber]["cites"] = getCites(arnumber)
self.data[arnumber]["keywords"] = index(arnumber)
self.data[arnumber]["layer"] = layer
