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
|
from neo4j import GraphDatabase
from typing import List, Dict, Optional
import asyncio
from datetime import datetime, timedelta
import json
class EnterpriseGraphService:
def __init__(self, uri: str, user: str, password: str):
self.driver = GraphDatabase.driver(uri, auth=(user, password))
def close(self):
self.driver.close()
async def bulk_import_social_network(self, users_data: List[Dict],
relationships_data: List[Dict]):
"""批量导入社交网络数据"""
def create_users_transaction(tx, batch_data):
query = """
UNWIND $batch as user
MERGE (u:User {id: user.id})
SET u.name = user.name,
u.email = user.email,
u.age = user.age,
u.interests = user.interests,
u.joinDate = datetime(user.joinDate)
RETURN COUNT(u) as created
"""
result = tx.run(query, batch=batch_data)
return result.single()["created"]
def create_relationships_transaction(tx, batch_data):
query = """
UNWIND $batch as rel
MATCH (u1:User {id: rel.from_id})
MATCH (u2:User {id: rel.to_id})
MERGE (u1)-[r:FOLLOWS]->(u2)
SET r.timestamp = datetime(rel.timestamp),
r.strength = rel.strength
RETURN COUNT(r) as created
"""
result = tx.run(query, batch=batch_data)
return result.single()["created"]
with self.driver.session() as session:
batch_size = 1000
for i in range(0, len(users_data), batch_size):
batch = users_data[i:i + batch_size]
users_created = session.execute_write(
create_users_transaction, batch
)
print(f"Created {users_created} users in batch {i // batch_size + 1}")
for i in range(0, len(relationships_data), batch_size):
batch = relationships_data[i:i + batch_size]
rels_created = session.execute_write(
create_relationships_transaction, batch
)
print(f"Created {rels_created} relationships in batch {i // batch_size + 1}")
def get_advanced_recommendations(self, user_id: str, limit: int = 10) -> List[Dict]:
"""获取高级推荐结果"""
def recommendation_transaction(tx, uid, lmt):
query = """
MATCH (target:User {id: $user_id})
// 基于共同好友的推荐
MATCH (target)-[:FOLLOWS]->(friend)-[:FOLLOWS]->(recommendation:User)
WHERE NOT (target)-[:FOLLOWS]->(recommendation)
AND recommendation <> target
WITH recommendation, COUNT(friend) as mutualFriends,
// 计算兴趣相似度
SIZE([interest IN target.interests
WHERE interest IN recommendation.interests]) as commonInterests
// 获取推荐用户的活跃度
OPTIONAL MATCH (recommendation)-[:POSTED]->(post:Post)
WHERE post.timestamp >= datetime() - duration({days: 30})
WITH recommendation, mutualFriends, commonInterests, COUNT(post) as recentPosts,
// 计算综合推荐分数
(mutualFriends * 2 + commonInterests * 3 + COUNT(post) * 0.1) as score
WHERE score > 1 // 过滤低分推荐
RETURN recommendation.id as userId,
recommendation.name as name,
recommendation.interests as interests,
mutualFriends,
commonInterests,
recentPosts,
score
ORDER BY score DESC
LIMIT $limit
"""
result = tx.run(query, user_id=uid, limit=lmt)
return [record.data() for record in result]
with self.driver.session() as session:
return session.execute_read(
recommendation_transaction, user_id, limit
)
def detect_fraud_patterns(self, account_id: str,
hours_back: int = 24) -> Dict:
"""检测欺诈模式"""
def fraud_detection_transaction(tx, acc_id, hours):
query = """
MATCH (account:Account {id: $account_id})
// 分析最近交易行为
MATCH (account)-[txn:TRANSACTION]->(target:Account)
WHERE txn.timestamp >= datetime() - duration({hours: $hours})
WITH account,
COUNT(txn) as transactionCount,
SUM(txn.amount) as totalAmount,
AVG(txn.amount) as avgAmount,
COLLECT(DISTINCT target.id) as uniqueTargets,
COLLECT(txn.amount) as amounts
// 计算统计指标
WITH account, transactionCount, totalAmount, avgAmount, uniqueTargets,
// 计算标准差
SQRT(REDUCE(s = 0.0, x IN amounts | s + (x - avgAmount)^2) / SIZE(amounts)) as stdDev
// 检查关联风险账户
OPTIONAL MATCH (account)-[:ASSOCIATED_WITH*1..2]-(riskAccount:Account)
WHERE riskAccount.status IN ['SUSPENDED', 'FLAGGED', 'FROZEN']
WITH account, transactionCount, totalAmount, avgAmount, uniqueTargets, stdDev,
COUNT(riskAccount) as riskAssociations
// 计算风险分数
WITH account, transactionCount, totalAmount, avgAmount, SIZE(uniqueTargets) as targetCount,
stdDev, riskAssociations,
// 复合风险评分算法
CASE
WHEN transactionCount > 100 THEN 5
WHEN transactionCount > 50 THEN 3
ELSE 1
END +
CASE
WHEN SIZE(uniqueTargets) > 30 THEN 4
WHEN SIZE(uniqueTargets) > 15 THEN 2
ELSE 0
END +
CASE
WHEN totalAmount > 50000 THEN 3
WHEN totalAmount > 20000 THEN 2
ELSE 0
END +
riskAssociations * 2 as riskScore
RETURN account.id as accountId,
transactionCount,
totalAmount,
targetCount,
riskAssociations,
riskScore,
CASE
WHEN riskScore >= 10 THEN 'HIGH'
WHEN riskScore >= 6 THEN 'MEDIUM'
WHEN riskScore >= 3 THEN 'LOW'
ELSE 'MINIMAL'
END as riskLevel
"""
result = tx.run(query, account_id=acc_id, hours=hours)
return result.single().data()
with self.driver.session() as session:
return session.execute_read(
fraud_detection_transaction, account_id, hours_back
)
async def main():
graph_service = EnterpriseGraphService(
"bolt://localhost:7687", "neo4j", "password"
)
try:
recommendations = graph_service.get_advanced_recommendations(
"user_12345", limit=5
)
print("推荐结果：", json.dumps(recommendations, indent=2))
fraud_analysis = graph_service.detect_fraud_patterns(
"account_67890", hours_back=48
)
print("欺诈分析：", json.dumps(fraud_analysis, indent=2))
finally:
graph_service.close()
if __name__ == "__main__":
asyncio.run(main())
|