基于Neo4j构建的外贸企业关系图谱做企业相似度查询「建议收藏」

'''初始化，用于连接到Neo4j'''
def __init__(self, data):
    self.data = data
    self.g = Graph(
        host="127.0.0.1",  # neo4j 搭载服务器的ip地址
        http_port=7474,  # neo4j 服务器监听的端口号
        user="neo4j",  # 数据库user name
        password="112233") # 密码

'''读取文件'''
def read_nodes(self):
    # 共5类节点
    enterprise = self.data['COMP_NAME_CH']  # 企业
    region = set(self.data['PROVINCE_CH'])  # 地区
    if (np.nan in region):
        region.remove(np.nan)
    country = []  # 出口国家
    for index, row in self.data.iterrows():
        for r in row['EXPORT_COUNTRY_MXT'].split(','):
            country.append(r)
    # 企业类型：1-manufacture-生产型、2-trader-贸易型（贸信通）3-服务型
    enterprise_type = ['生产型', '贸易型', '服务型']  # 企业类型
    legal_representative = self.data['LEGAL_REPRESENTATIVE']  # 法人代表

    # 构建节点实体关系
    rels_region = []  # 企业－地区关系 locate
    rels_country = []  # 企业－出口国家关系 export
    rels_type = []  # 企业－企业类型关系 type
    # rels_product = [] # 企业－产品关系 product
    rels_legal = []  # 企业－法人代表关系 legal
    for index, row in self.data.iterrows():
        if (row['PROVINCE_CH'] is not np.nan):
            rels_region.append([row['COMP_NAME_CH'], row['PROVINCE_CH']])
        for r in row['EXPORT_COUNTRY_MXT'].split(','):
            # 一个企业有多个出口国家
            rels_country.append([row['COMP_NAME_CH'], r])
        rels_type.append([row['COMP_NAME_CH'], '生产型' if row['COMP_TYPE'] == 1\
            else ('服务型' if row['COMP_TYPE'] == 2 else '贸易型')])
        rels_legal.append([row['COMP_NAME_CH'], row['LEGAL_REPRESENTATIVE']])

    return set(enterprise), set(region), set(country), set(enterprise_type), set(legal_representative), \
           rels_region, rels_country, rels_type, rels_legal

'''建立单标签节点'''
def create_node(self, label, nodes):
    count = 0
    for node_name in nodes:
        node = Node(label, name=node_name)
        self.g.create(node)
        count += 1
        print(count, len(nodes))
    return

'''创建知识图谱外贸企业的节点'''
def create_enterprise_nodes(self):
    count = 0
    for index, row in self.data.iterrows():
        node = Node("Enterprise", name=row['COMP_NAME_CH'], credit_code=row['CREDIT_CODE'],
                    setup_time=row['SETUP_TIME'], address=row['ADDRESS_CH'],
                    captial=str(row['REG_CAPITAL']) + '万人民币')
        self.g.create(node)
        count += 1
        print(count)
    return

'''创建实体关联边'''
def create_relationship(self, start_node, end_node, edges, rel_type, rel_name):
    count = 0
    # 去重处理
    set_edges = []
    for edge in edges:
        set_edges.append('###'.join(edge))
    all = len(set(set_edges))
    for edge in set(set_edges):
        edge = edge.split('###')
        p = edge[0]
        q = edge[1]
        query = "match(p:%s),(q:%s) where p.name='%s'and q.name='%s' create (p)-[rel:%s{name:'%s'}]->(q)" % (
            start_node, end_node, p, q, rel_type, rel_name)
        try:
            self.g.run(query)
            count += 1
            print(rel_type, count, all)
        except Exception as e:
            print(e)
    return

'''创建知识图谱实体节点类型schema'''
def create_graphnodes(self):
    # 获取所有节点和关系
    Enterprises, Regions, Countries, Enterprise_types, Legal_representatives, \
    rels_region, rels_country, rels_type, rels_legal = self.read_nodes()
    # 创建图数据库节点
    self.create_enterprise_nodes() # 企业
    self.create_node('Region', Regions) # 地区
    print('地区：' + str(len(Regions)))
    self.create_node('Country', Countries) # 出口国家
    print('出口国家：' + str(len(Countries)))
    self.create_node('Type', Enterprise_types)  #企业类型
    print('企业类型：' + str(len(Enterprise_types)))
    # 暂不需要使用该节点和关系
    # self.create_node('Legal', Legal_representatives) # 法人代表
    # print('法人代表：' + len(Legal_representatives))
    return

'''创建实体关系边'''
def create_graphrels(self):
    # 获取所有关系组
    Enterprises, Regions, Countries, Enterprise_types, Legal_representatives, \
    rels_region, rels_country, rels_type, rels_legal = self.read_nodes()
    self.create_relationship('Enterprise', 'Region', rels_region, 'locate', '所在地区')
    self.create_relationship('Enterprise', 'Country', rels_country, 'export', '出口')
    self.create_relationship('Enterprise', 'Type', rels_type, 'type', '类型')
    # 暂不需要导入该关系
    # self.create_relationship('Enterprise', 'Legal', rels_legal, 'legal', '法人')

if __name__ == '__main__':
    # 获取当前路径，并转换为正确格式
    cur_dir = '/'.join(os.path.abspath(__file__).split('\\')[:-1])
    data_path = cur_dir + '/TB_ENTERPRISEINFO_FUSE_BAK.csv'
    print('read_csv from:' + data_path)
    data = pd.read_csv(data_path)
    # 创建实例
    handler = EnterpriseGragh(data)
    # 构建企业图谱的节点和关系
    handler.create_graphnodes()
    handler.create_graphrels()

match p=(n:Enterprise{name:'陕西和沃进出口有限公司'})-[*2..3]-() return p limit 20

MATCH (n:Enterprise{name:'陕西和沃进出口有限公司'})-[:export]->(c:Country)<-[:export]-(other:Enterprise)
with n,other,count(c) as intersection,collect(c.name) as collection
match (n)-[:export]->(nc:Country)
with n,other,intersection,collection,collect(nc.name) as s1
match (other)-[:export]->(oc:Country)
with n,other,intersection,collection,s1,collect(oc.name) as s2
with n,other,intersection,s1,s2
with n,other,intersection,s1+filter(x IN s2 where not x IN s1) as uni,s1,s2
return n.name,other.name,s1,s2,((1.0*intersection)/SIZE(uni)) as jaccard
order by jaccard DESC
limit 20

MATCH (n:Enterprise) where n.name='陕西和沃进出口有限公司'
match (n)-[:type]->(t:Type)<-[:type]-(other:Enterprise)
with n,other,count(t) as tn
optional match (n)-[:locate]->(r:Region)<-[:locate]-(other)
with n,other,tn,count(r) as rn
optional match (n)-[:export]->(c:Country)<-[:export]-(other)
with n,other,tn,rn,count(c) as cn
return other.name as 推荐企业,tn as 相同企业类型,rn as 相同地区,cn as 相同出口国家,(3*tn)+(3*rn)+(1*cn) as score
ORDER BY score DESC
limit 100