-
Notifications
You must be signed in to change notification settings - Fork 0
/
DBLP_Scholar_EA.py
229 lines (207 loc) · 9.03 KB
/
DBLP_Scholar_EA.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
import numpy as np
import pandas as pd
import editdistance
# import gensim
from neo4j import GraphDatabase, basic_auth,kerberos_auth,custom_auth,TRUST_ALL_CERTIFICATES
'''
需要先构建好图谱才能进行融合操作
'''
#链接数据库
driver = GraphDatabase.driver("neo4j://localhost:7687", auth=basic_auth("neo4j","admin"), encrypted=False)
session = driver.session()
csv_data = pd.read_csv("./data/DBLP-Scholar_perfectMapping.csv")
csv_data['idDBLP'] = csv_data['idDBLP'].map(lambda x: str(x))
# #加载预训练模型
# word2vec = gensim.models.KeyedVectors.load_word2vec_format("./data/GoogleNews-vectors-negative300.bin.gz", binary=True)
#读取Neo4j的值
data_DBLP_paper = session.run("MATCH (DBLP_paper:DBLP_paper) return DBLP_paper.DBLP_title,DBLP_paper.DBLP_id,DBLP_paper.year ")
data_Scholar_paper = session.run("MATCH (Scholar_paper:Scholar_paper) return Scholar_paper.Scholar_title,Scholar_paper.Scholar_id,Scholar_paper.year,Scholar_paper.remark ")
#初始化
dlists = []
dlists_id = []
dlists_year = []
dlists_t = []
slists = []
slists_id = []
slists_year = []
slists_t = []
slists_remark = []
zero = 0
#把DBLP的作者结点导入dlist
for d in data_DBLP_paper:
bs = str(d[0])
bs_id = str(d[1])
bs_year = str(d[2])
dlists.append(bs)
dlists_id.append(bs_id)
dlists_year.append(bs_year)
#把Scholar的作者结点导入slist
for d in data_Scholar_paper:
bs = str(d[0])
bs_id = str(d[1])
bs_year = str(d[2])
bs_remark = str(d[3])
slists.append(bs)
slists_id.append(bs_id)
slists_year.append(bs_year)
slists_remark.append(bs_remark)
#编辑距离函数
def edit_distance(word1, word2):
# len1 = len(word1)
# len2 = len(word2)
# dp = np.zeros((len1 + 1, len2 + 1))
# for i in range(len1 + 1):
# dp[i][0] = i
# for j in range(len2 + 1):
# dp[0][j] = j
# for i in range(1, len1 + 1):
# for j in range(1, len2 + 1):
# delta = 0 if word1[i - 1] == word2[j - 1] else 1
# dp[i][j] = min(dp[i-1][j-1] + delta, min(dp[i-1][j] + 1, dp[i][j-1] + 1))
# return dp[len1][len2]
distance = editdistance.eval(word1,word2)
return distance
#杰卡德函数
def Jaccrad(terms_model,reference):
grams_reference = set(reference)
grams_model = set(terms_model)
temp = 0
for i in grams_reference:
if i in grams_model:
temp = temp + 1
fenmu = len(grams_model) + len(grams_reference) - temp
jaccard_coefficient = float(temp/fenmu)
return jaccard_coefficient
#查询DBLP文章对应的作者
def DBLP_author(a):
DBLP_paper_author = session.run("MATCH p=(m)-[r:DBLP_publish]->(DBLP_paper{DBLP_id:'%s'}) return m.name" % (a))
for d in DBLP_paper_author:
bs = str(d[0])
dlists_t.append(bs)
return dlists_t
#查询Scholar文章对应的作者
def Scholar_author(a):
Scholar_paper_author = session.run("MATCH p=(m)-[r:Scholar_publish]->(Scholar_paper{Scholar_id:'%s'}) return m.name" % (a))
for d in Scholar_paper_author:
bs = str(d[0])
slists_t.append(bs)
return slists_t
#查询DBLP的会议信息
def DBLP_Venue(a):
DBLP_Venue = session.run("MATCH p=(DBLP_paper{DBLP_id:'%s'})-[r:DBLP_included]->(m) return m.DBLP_V_ID" % (a))
return DBLP_Venue
#查询Scholar的会议的信息
def Scholar_Venue(a):
Scholar_venue = session.run("MATCH p=(Scholar_paper{DBLP2_id:'%s'})-[r:Scholar_included]->(m) return m.Scholar_V_ID" % (a))
return Scholar_venue
#分词函数
def spilt_sentence(sentence):
words = sentence.spilt()
return words
TP = 0.001
FP = 0.001
FN = 5347
step = 0
tag = True
Flag = True #查询最佳匹配是否正确标记
#把数据拿出来两两比较
for i in range(len(dlists)):
for j in range(len(slists)):
a = dlists[i]
b = slists[j]
DBLP_id = dlists_id[i]
Scholar_id = slists_id[j]
DBLP_year = int(dlists_year[i])
if slists_year[j] != 'null':
temp = float(slists_year[j])
Scholar_year = int(temp)
#计算相似度
jacd = Jaccrad(a,b)
std = edit_distance(a,b)/max(len(a),len(b))
edit = 1-std
#计算平均值
huizon = (jacd+edit)/2
#判断
if 0.70 < huizon < 1.1:
#查找在匹配集中的数据
data = csv_data.loc[csv_data['idDBLP'] == DBLP_id, 'idScholar']
#判断标记
tag = True
#判断进行匹配
if Scholar_year != 'null':
if DBLP_year == Scholar_year:
tag = True
else:
tag = False
#循环查询数组判断是否正确
if data.shape != (0,):
for m in range(len(data)):
data_t = str(data.values[m])
if data_t == Scholar_id:
Flag = True
break
else:
Flag = False
if Flag and tag:
TP+=1
#生成匹配成功的表格
csv_data.loc[csv_data['idDBLP'] == DBLP_id, 'Scholar_id'] = 'CORRECT!!!!!!'
#print("<DBLP_name,Scholar_name>-<%s,%s>-<%s,%s>|平均相似度: %f " % (dlists[i], slists[j], dlists_id[i],slists_id[j],huizon))
#----------------------------------融合功能------------------------------------------------
# #文章结点融合
# try:
# session.run(" MATCH (a1:ACM_paper {ACM_id:'%s'}),(a2:DBLP2_paper {DBLP2_id:'%s'}) WITH head(collect([a1,a2])) as nodes CALL apoc.refactor.mergeNodes(nodes,{properties:'combine', mergeRels:true}) yield node return nodes" % (ACM_id,DBLP2_id))
# except:
# #文章对应的会议结点融合
# A_venue_ID = ACM_Venue(ACM_id)
# D_venue_ID = DBLP2_Venue(DBLP2_id)
# session.run("MATCH (a1:ACM_venue {venue:'%s'}),(a2:DBLP2_venue {venue:'%s'}) WITH head(collect([a1,a2])) as nodes CALL apoc.refactor.mergeNodes(nodes,{properties:'combine', mergeRels:true}) yield node return nodes" % (A_venue_ID,D_venue_ID))
# #查询文章对应的作者
# alists_t = ACM_author(ACM_id)
# dlists_t = DBLP2_author(DBLP2_id)
# print(alists_t,dlists_t)
# #文章对应的作者的结点融合,在文章对应的数据集中进行匹配，大于0.5的就进行融合
# for n in range(len(alists_t)):
# for m in range(len(dlists_t)):
# aa = alists_t[n]
# bb = dlists_t[m]
# #计算作者结点的相似度
# jacd_t = Jaccrad(aa,bb)
# std_t = edit_distance(aa,bb)/max(len(aa),len(bb))
# edit_t = 1-std_t
# huizon_t = (jacd_t+edit_t)/2
# #平均相似度大于0.69就进行融合
# if huizon_t > 0.69:
# print('-------------------------即将做融合的作者结点')
# print(huizon_t)
# print(aa,bb)
# try:
# session.run("MATCH (a1:ACM_author {name:'%s'}),(a2:DBLP2_author {name:'%s'}) WITH head(collect([a1,a2])) as nodes CALL apoc.refactor.mergeNodes(nodes,{properties:'combine', mergeRels:true}) yield node return nodes " % (aa,bb))
# except:
# pass
# #清空list
# alists_t.clear()
# dlists_t.clear()
#--------------------------------融合功能-----------------------------------------
else:
FP+=1
print("<DBLP_name,Scholar_name>-<%s,%s>|平均相似度: %f " % (a, b, huizon))
print('------------------------不正确嗷！')
#print(ACM_id, DBLP2_id)
csv_data.loc[csv_data['idDBLP'] == DBLP_id, 'idScholar'] = 'ERROR!!!!'
csv_data.loc[csv_data['idDBLP'] == DBLP_id, 'DBLP_name'] = a
csv_data.loc[csv_data['idDBLP'] == DBLP_id, 'Scholar_name'] = b
else:
data = 'xxxx'
#每一步的情况
step = step + 1
if step % 1000000 == 0:
Perecision_t = TP/(TP+FP)
print('------------------------------------第 %d 精确度是 %f ：' % (step,Perecision_t))
Perecision = TP/(TP+FP)
FN = FN - TP
Recall = TP/(TP+FN)
print('最终精确度：', Perecision)
print('最终召回率：', Recall)
#生成匹配错误日志
csv_data.to_csv('DBLP_Scholar_matching_excel.csv', encoding='utf-8', index=0)