neo4j安装及配置，绘制人物关系图谱

1. 先升级pip，安装py2neo

pip install py2neo==2021.0.1

2. 依赖 jdk1.8， neo4j 3.xx；
   或者jdk18，neo4j 4.x，5.x；
   官网下载了neo4j4.x,5.x 因为jdk版本原因都不行，最终选择了neo4j3.xx

• 下载：https://neo4j.com/artifact.php?name=neo4j-community-3.5.31-windows.zip
  https://neo4j.com/download-center
  安装，配置环境变量，启动，浏览器打开

neo4就 5.x，4.x都不行，与java1.8不兼容，最后安了3.5.31 ok，忽略文件夹名称为了不用重复配置环境变量 NEO4j_HOME及path

终于安装好了✌ ✌ ✌

1. 效果图

人物图谱效果图如下：

群山回唱人物关系图谱 效果图如下：

2. 源码

# 生成人物关系图谱import pandas as pd
from py2neo import Graph, Node, Relationship# 字段映射字典
mapping_dict = {"发生时间": "occurrence_time","发布时间": "release_time","发生地点": "scene","死亡人数": "death_toll","受伤人数": "injury_toll","直接损失": "direct_loss","事故主体单位": "accident_unit","灾害事故类型": "accident_type","防范与整改": "prevention_rectification","处理建议": "treatment_recommendations","应急措施": "emergency_measure","直接原因": "immediate_cause","严重程度": "order_severity","应急响应": "emergency_response"}
# 读取文件
storage_df = pd.read_csv('files/new.csv', encoding='utf-8')
# 获取列标签
columns_list = storage_df.columns.tolist()
# 获取数据条数
nums = len(storage_df['title'])
# 删除title列标签
columns_list.remove('title')
# 逆转映射字典
new_dic = dict(zip(mapping_dict.values(), mapping_dict.keys()))# 连接数据库，输入个人配置
graph = Graph("http://localhost:7474/", username='neo4j', password='neo4j123456', run="sub")
# 清空全部数据
graph.delete_all()
# 开启一个新的事务
graph.begin()for i in range(nums):data_dict = {}title = storage_df['title'][i]for columns in columns_list:if str(storage_df[columns][i]) != 'nan':data_dict[columns] = storage_df[columns][i]# 创建事件节点（主节点），节点属性node1 = Node("case", name=title, **data_dict)graph.create(node1, 'case', 'name')# 删除id列data_dict.pop('id')# 创建关系与副节点for key, value in data_dict.items():# 创建副节点node2 = Node(key, name=value)graph.create(node2, key, 'name')# 创建关系rel = Relationship(node1, new_dic[key], node2, type=key)print(new_dic[key])graph.create(rel)

2.1 群山会唱人物关系图谱

# python plot_qshc_rwtp.py
# 对群山回唱人物关系图谱进行绘制from py2neo import Graph, Node, Relationship# 连接数据库，输入个人配置
graph = Graph("http://localhost:7474/", username='xx', password='xx', run="sub")
# 清空全部数据
graph.delete_all()
# 开启一个新的事务
graph.begin()dict_nodes = {}
with open("qshc.txt", 'r', encoding='utf-8', errors='ignore') as f:data = f.readlines()print(data)for i, line in enumerate(data):print(i, line.replace("\n", ""))node_a = line.split(",")[0]node_b = line.split(",")[1]relation = line.split(",")[2]# 创建副节点node1 = Nonenode2 = Noneif (dict_nodes.__contains__(node_a)):node1 = dict_nodes.get(node_a)else:node1 = Node(node_a, name=node_a)graph.create(node1)dict_nodes[node_a] = node1if (dict_nodes.__contains__(node_b)):node2 = dict_nodes.get(node_b)else:node2 = Node(node_b, name=node_b)graph.create(node2)dict_nodes[node_b] = node2# 创建关系rel = Relationship(node1, relation, node2, type='key')graph.create(rel)

参考

• https://blog.csdn.net/qq_52159080/article/details/124018875
• https://blog.csdn.net/qq_52159080/article/details/124018875