push

project5/基于医疗知识图谱的问答系统/student_code/.vscode/launch.json

+{
+{
+    // 使用 IntelliSense 了解相关属性。 
+    // 悬停以查看现有属性的描述。
+    // 欲了解更多信息，请访问: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "Python: Current File",
+            "type": "python",
+            "request": "launch",
+            "program": "${file}",
+            "console": "integratedTerminal",
+            "justMyCode": false
+        }
+    ]
+}
\ No newline at end of file

project5/基于医疗知识图谱的问答系统/student_code/aa.py

+from module.semantic_parser import build_search_tree
+from module.semantic_parser import build_search_tree
+from config import *
+
+print('开始训练实体搜索树...')
+
+build_search_tree(entity_corpus_path, entity_searcher_save_path)
+
+print('实体搜索树训练成功...')
\ No newline at end of file

project5/基于医疗知识图谱的问答系统/student_code/config.py

+# 知识语料路径
+# 知识语料路径
+entity_corpus_path = './data/knowledge/'
+
+# 实体搜索器存储路径
+entity_searcher_save_path = './checkpoints/entity_searcher/search_tree.pkl'
+
+# 实体搜索器加载路径
+entity_searcher_load_path = './checkpoints/entity_searcher/search_tree.pkl'
+
+# 分类器语料路径
+classifier_corpus_path = './data/classifier/chat.train'
+
+# 分类器模型存储路径
+classifier_save_path = './checkpoints/classifier/model.bin'
+
+# 分类器模型加载路径
+classifier_load_path = './checkpoints/classifier/model.bin'
+
+# 闲聊回复语料库
+chat_responses = {
+    'greet': [
+        'hello，我是小A，小哥哥小姐姐有关于股票的问题可以问我哦',
+        '你好，我是小A，输入股票名称或者代码查看详细信息哦',
+        '你好，我是小A，可以问我股票相关的问题哦'
+    ],
+    'goodbye': [
+        '再见',
+        '不要走，继续聊会呗',
+        '拜拜喽，别忘了给个小红心啊',
+    ],
+    'bot': [
+        '没错，我就是集美貌与才智于一身的小A',
+        '小A就是我，我就是小A'
+    ],
+    'safe': [
+        '祝你吃嘛嘛香，身体倍棒',
+        '人是铁饭是钢，一顿不吃得慌'
+    ]
+}
+
+# 问题类型
+question_types = {
+    'concept':
+        ['概念'],
+    'holder':
+        ['股东', '控制', '控股', '持有'],
+    'industry':
+        ['行业'],
+}
+
+# 存储对话历史中上一次涉及的问题类型和实体
+contexts = {
+    'ques_types': None,
+    'entities': None
+}

project5/基于医疗知识图谱的问答系统/student_code/data/classifier/chat.train

+__label__greet 打扰 一下
+__label__greet 打扰 一下
+__label__bot what   your   name ?
+__label__bot 你 是 机器人 吗
+__label__bot 你 的 名字 ？
+__label__greet 你好
+__label__bot 你 叫 什么 啊
+__label__bot nishishui
+__label__greet 打扰 了
+__label__greet 晚上 好
+__label__greet hey
+__label__goodbye 忙 事情 去 了
+__label__bot 你 叫 什么 名字 啊
+__label__bot 你 到底 是 哪个
+__label__greet 有 人工 服务 吗
+__label__greet hello
+__label__greet 有个 问题 请教
+__label__bot 怎么 称呼 你 啊 ？
+__label__goodbye 回见
+__label__bot 谁 呢
+__label__bot 机器人 ？
+__label__bot 你 的 名字 是 什么
+__label__greet 您好
+__label__greet good   afternoon
+__label__goodbye 再见 啦 ， 下次 聊
+__label__bot 你 是 机器人 啊
+__label__greet morning
+__label__bot 你是谁
+__label__bot 你 是 人类 吗 ？
+__label__greet nihao
+__label__greet 下午 好 啊
+__label__bot 你 是 人 吗 ？
+__label__goodbye bye ~
+__label__bot 报上名来
+__label__greet 哈 喽
+__label__goodbye 拜拜 咯
+__label__greet 嘿嘿嘿
+__label__goodbye bye
+__label__greet hey   men
+__label__bot 你 是 谁 呢
+__label__goodbye 先 这样 吧
+__label__goodbye goodbye
+__label__greet 嗨
+__label__goodbye 拜拜
+__label__greet good   evening
+__label__goodbye 再见
+__label__bot 你 是 谁
+__label__greet 现在 营业 吗
+__label__goodbye 不聊 了
+__label__bot 怎么 称呼
+__label__bot 谁 在 呢
+__label__bot 谁 啊
+__label__greet hi
+__label__goodbye 我 去 忙 了
+__label__bot 你 是 哪位
+__label__greet what   up
+__label__goodbye 下次 见
+__label__greet good   morning
+__label__greet 有人 吗
+__label__greet good   night
+__label__greet 你好 啊
+__label__greet 你好 呀
+__label__bot 你 是 哪个 ？ ？
+__label__goodbye 再见 咯
+__label__greet 可以 问 你 问题 吗
+__label__greet 早上好 呀
+__label__bot who   are   you
+__label__greet 你 有 时间 吗
+__label__bot 来者 何人
+__label__greet 很 高心 见到 你
+__label__bot 你 叫 什么
+__label__bot 你 是 谁 啊 ？
+__label__bot 你 是 ？
+__label__greet 嘿
+__label__greet hai
+__label__bot 你 是 人 还是 机器人
\ No newline at end of file

project5/基于医疗知识图谱的问答系统/student_code/data/knowledge/概念信息.csv

+code,name,src
+code,name,src
+TS0,密集调研,ts
+TS1,南北船合并,ts
+TS2,5G,ts
+TS3,机场,ts
+TS4,高价股,ts
+TS5,烧碱,ts
+TS6,AH溢价股,ts
+TS7,保险,ts
+TS8,PVC,ts
+TS9,啤酒,ts
+TS10,火电,ts
+TS11,银行,ts
+TS12,碳纤维,ts
+TS13,安邦系,ts
+TS14,特高压,ts
+TS16,光通信,ts
+TS17,草甘膦,ts
+TS18,高速公路,ts
+TS19,有色-铝,ts
+TS20,有色-锆,ts
+TS21,主题公园,ts
+TS22,白马股,ts
+TS23,雄安-地产,ts
+TS24,OLED,ts
+TS25,MSCI,ts
+TS26,健康中国,ts
+TS27,养老金持股,ts
+TS28,水电,ts
+TS29,环氧丙烷,ts
+TS30,PCB板,ts
+TS31,煤炭,ts
+TS32,中字头,ts
+TS33,血制品,ts
+TS34,钢铁,ts
+TS35,油服,ts
+TS36,内蒙古,ts
+TS37,券商,ts
+TS38,有色-锌,ts
+TS39,人民币升值受益,ts
+TS40,阿里巴巴概念股,ts
+TS41,有色-锑,ts
+TS42,海工装备,ts
+TS43,新能源整车,ts
+TS44,涤纶,ts
+TS45,振兴东北,ts
+TS46,有色-镁,ts
+TS47,油气改革,ts
+TS48,一带一路,ts
+TS49,新进指数成份股,ts
+TS50,玻纤,ts
+TS51,大消费,ts
+TS52,被动元件,ts
+TS53,北汽新能源,ts
+TS54,可燃冰,ts
+TS55,东盟博览会,ts
+TS56,电改,ts
+TS57,石墨电极,ts
+TS58,北京城市规划,ts
+TS59,特斯拉,ts
+TS60,赛马,ts
+TS61,航运,ts
+TS62,金改,ts
+TS63,超级高铁,ts
+TS64,农药,ts
+TS65,汽车整车,ts
+TS66,次新股,ts
+TS67,安徽国企改革,ts
+TS68,水务,ts
+TS69,混改,ts
+TS70,PD-1抑制剂,ts
+TS71,二胎,ts
+TS72,固废处理,ts
+TS73,医药商业,ts
+TS74,教育,ts
+TS75,军工混改,ts
+TS76,海底捞概念股,ts
+TS78,硅锰,ts
+TS79,高送转-预期,ts
+TS80,央企改革,ts
+TS81,有机硅,ts
+TS82,民营医院,ts
+TS83,华为产业链,ts
+TS84,蔚来汽车概念股,ts
+TS85,甲醇,ts
+TS86,军工研究所,ts
+TS87,港口,ts
+TS88,社保重仓,ts
+TS89,美团概念股,ts
+TS90,纯碱,ts
+TS91,转基因,ts
+TS92,旅游,ts
+TS93,福建自贸区,ts
+TS94,核电,ts
+TS95,乡村振兴,ts
+TS96,调味品,ts
+TS97,大飞机,ts
+TS98,回购,ts
+TS99,通用航空,ts
+TS100,白酒,ts
+TS101,汽车销售,ts
+TS102,重庆国企改革,ts
+TS103,制冷剂,ts
+TS104,航母,ts
+TS105,新零售,ts
+TS106,饲料,ts
+TS107,空铁WiFi,ts
+TS108,有色金属,ts
+TS109,养猪,ts
+TS110,芬太尼概念,ts
+TS111,业绩预增,ts
+TS112,参股券商,ts
+TS113,园林,ts
+TS114,宝能系,ts
+TS115,破净股,ts
+TS116,饮料,ts
+TS117,风电,ts
+TS118,粤港澳大湾区,ts
+TS119,棉花,ts
+TS120,中科院系,ts
+TS121,页岩气,ts
+TS122,重卡,ts
+TS123,供应链金融,ts
+TS124,保健品,ts
+TS125,锦纶（尼龙）,ts
+TS126,染料,ts
+TS127,军工,ts
+TS128,纺织服装,ts
+TS129,天然气,ts
+TS130,军民融合,ts
+TS131,石油化工,ts
+TS132,浙江国企改革,ts
+TS133,无人机,ts
+TS134,有色-镍,ts
+TS135,养鸡,ts
+TS136,兜底式增持,ts
+TS137,有色-钴,ts
+TS138,汽车零部件,ts
+TS139,有色-钼,ts
+TS140,炭黑,ts
+TS141,大基建,ts
+TS142,物联网,ts
+TS143,百度概念股,ts
+TS144,筹码集中,ts
+TS145,宁德时代概念股,ts
+TS146,光伏,ts
+TS147,信托,ts
+TS148,储能,ts
+TS149,装配式建筑,ts
+TS150,煤化工,ts
+TS151,脑科学,ts
+TS152,食品,ts
+TS153,无人驾驶,ts
+TS154,债转股-AMC,ts
+TS155,湖南国企改革,ts
+TS156,新疆,ts
+TS157,雄安-园林环保,ts
+TS158,雄安-金融,ts
+TS159,中韩自贸区,ts
+TS160,恒大概念股,ts
+TS161,智慧物流,ts
+TS162,谷歌概念股,ts
+TS163,雄安-智慧城市,ts
+TS164,化肥,ts
+TS165,特色小镇,ts
+TS166,养老产业,ts
+TS167,厨卫家电,ts
+TS168,农垦,ts
+TS169,农机,ts
+TS170,定增破发,ts
+TS171,稀土磁材,ts
+TS172,休闲食品,ts
+TS173,小家电,ts
+TS174,有色-钨,ts
+TS175,工业自动化,ts
+TS176,智能仪表,ts
+TS177,雄安-建材,ts
+TS178,乳业（奶粉）,ts
+TS179,山西国企改革,ts
+TS180,水泥,ts
+TS181,地下管廊,ts
+TS182,轮胎,ts
+TS183,铁总混改,ts
+TS184,机械,ts
+TS185,环杭州湾大湾区,ts
+TS186,高质押率,ts
+TS187,雄安-公共服务,ts
+TS188,河南自贸区,ts
+TS189,山东国企改革,ts
+TS190,医药,ts
+TS191,太赫兹,ts
+TS192,万达系,ts
+TS193,白色家电,ts
+TS194,白糖,ts
+TS195,水利,ts
+TS196,中药,ts
+TS197,智慧停车,ts
+TS198,大气治理,ts
+TS199,房地产,ts
+TS200,创新药,ts
+TS201,海绵城市,ts
+TS202,北斗导航,ts
+TS203,铁矿石,ts
+TS204,网络安全,ts
+TS205,家电,ts
+TS206,土地流转,ts
+TS207,化学药,ts
+TS208,锂电池,ts
+TS209,海南自由贸易港,ts
+TS210,PPP,ts
+TS211,小米概念股,ts
+TS212,疫苗,ts
+TS213,林业,ts
+TS214,磷化工,ts
+TS215,燃料电池,ts
+TS216,量子通信,ts
+TS218,环保,ts
+TS219,3D打印,ts
+TS220,数字中国,ts
+TS221,智能制造,ts
+TS222,创业板人气股,ts
+TS223,直播/短视频,ts
+TS224,冷链,ts
+TS225,雄安新区,ts
+TS226,新疆国企改革,ts
+TS227,互联网金融,ts
+TS228,有色-钒,ts
+TS229,无线充电,ts
+TS230,维生素,ts
+TS231,上海国企改革,ts
+TS232,新能源汽车,ts
+TS233,基因测序,ts
+TS234,雄安-工程基建,ts
+TS235,京津冀,ts
+TS236,快递物流,ts
+TS237,医疗器械,ts
+TS238,移动支付,ts
+TS239,大数据,ts
+TS240,油品升级,ts
+TS241,低价股,ts
+TS242,醋酸,ts
+TS243,流感,ts
+TS244,药品信息化追溯,ts
+TS245,有色-铜,ts
+TS246,云计算,ts
+TS247,进口博览会,ts
+TS248,人民币贬值受益,ts
+TS249,举牌股,ts
+TS250,住房租赁,ts
+TS251,高送转,ts
+TS252,自贸区,ts
+TS253,耐火材料,ts
+TS254,人工智能,ts
+TS255,草铵膦,ts
+TS256,造纸,ts
+TS257,石墨烯,ts
+TS258,互联网医疗,ts
+TS259,眼科,ts
+TS260,污水处理,ts
+TS261,区块链,ts
+TS262,玻璃,ts
+TS263,足球,ts
+TS264,工业互联网,ts
+TS265,指纹识别,ts
+TS266,半导体,ts
+TS267,3D感应,ts
+TS268,无人商店,ts
+TS269,智慧城市,ts
+TS270,农业种植,ts
+TS271,3D玻璃,ts
+TS272,印染,ts
+TS273,西藏,ts
+TS274,雄安-周边,ts
+TS275,跨境电商,ts
+TS276,粘胶短纤,ts
+TS277,国产软件,ts
+TS278,智能音箱,ts
+TS279,天津国企改革,ts
+TS280,自由贸易港,ts
+TS281,广西,ts
+TS282,股权转让,ts
+TS283,电竞,ts
+TS284,彩票,ts
+TS285,钛白粉,ts
+TS286,复牌股,ts
+TS287,精准医疗,ts
+TS288,京东金融概念股,ts
+TS289,民爆,ts
+TS290,VR&AR,ts
+TS291,腾讯概念股,ts
+TS292,家具家居,ts
+TS293,苹果产业链,ts
+TS294,高校系,ts
+TS295,国产芯片,ts
+TS296,滴滴出行概念股,ts
+TS297,共享经济,ts
+TS298,期货概念,ts
+TS299,黑色家电,ts
+TS300,地热,ts
+TS301,天津自贸区,ts
+TS302,迈瑞医疗概念股,ts
+TS303,医疗信息化,ts
+TS304,动物保健,ts
+TS305,电子发票,ts
+TS306,股权争夺,ts
+TS307,国产操作系统,ts
+TS308,高铁轨交,ts
+TS309,体外诊断,ts
+TS310,机器视觉,ts
+TS311,国资入股,ts
+TS312,包装印刷,ts
+TS313,智慧安防,ts
+TS314,资产重组,ts
+TS315,独角兽,ts
+TS316,年报季报,ts
+TS317,机器人,ts
+TS318,LED,ts
+TS319,碳交易,ts
+TS320,手势识别,ts
+TS321,ST股,ts
+TS322,蚂蚁金服概念股,ts
+TS323,充电桩,ts
+TS324,游戏,ts
+TS325,装修装饰,ts
+TS326,有色-钛,ts
+TS327,手游,ts
+TS328,虹膜识别,ts
+TS329,广电,ts
+TS330,深圳本地股/深汕合作,ts
+TS331,化妆品,ts
+TS332,体育产业,ts
+TS333,影视,ts
+TS334,壳公司,ts
+TS335,黄金,ts
+TS336,海南,ts
+TS337,知识产权,ts
+TS338,燃料乙醇,ts
+TS339,ST摘帽,ts
+TS340,破发次新,ts
+TS341,上海自由贸易港,ts
+TS342,民航,ts
+TS343,宁夏,ts
+TS344,人脸识别,ts
+TS345,强势人气股,ts
+TS346,橡胶,ts
+TS347,动漫,ts
+TS348,苹果期货,ts
+TS349,无人银行,ts
+TS350,富勒烯,ts
+TS351,科创板概念股,ts
+TS352,创投,ts
+TS353,电子车牌,ts
+TS354,雄安-设计,ts
+TS355,工业大麻,ts
+TS356,透明工厂,ts
+TS357,人造肉,ts
+TS358,垃圾分类,ts
+TS359,科创板,ts
+TS360,科创板对标,ts
+TS361,新型病毒,ts
\ No newline at end of file

project5/基于医疗知识图谱的问答系统/student_code/main.py

+from module.classifier import Classifier
+from module.classifier import Classifier
+from module.semantic_parser import SemanticParser
+from module.graph_matcher import GraphMatcher
+from config import classifier_load_path, entity_searcher_load_path, chat_responses, question_types
+from random import choice
+
+# 加载分类器
+classifier = Classifier(classifier_load_path)
+
+# 加载语义解析器，预测问题类型和涉及的实体
+semantic_parser = SemanticParser(entity_searcher_load_path, question_types)
+
+# 加载图数据库查询
+graph_matcher = GraphMatcher()
+
+
+while True:
+    query = input('用户: ')
+    if query == 'stop':
+        break
+    else:
+        # 预测 label 和概率
+        query_intent_label, query_intent_prob = classifier.predict(query)
+
+        # 闲聊
+        if query_intent_prob > 0.9:
+            response = choice(chat_responses[query_intent_label])
+        # 知识问答
+        else:
+            semantics = semantic_parser.predict(query)
+            if len(semantics['ques_types']) > 0 and len(semantics['entities']) > 0:
+                response = graph_matcher.predict(semantics)
+            else:
+                response = choice(chat_responses['safe'])
+        print(f'机器人: {response}')
+
+        if query_intent_label == 'goodbye':
+            break

project5/基于医疗知识图谱的问答系统/student_code/module/classifier.py

+import fasttext
+import fasttext
+import jieba
+from config import classifier_corpus_path, classifier_save_path
+
+def train_classifier(input_file_path, model_save_path):
+    """训练分类模型"""
+
+    # 基于 fasttext api 实现模型训练
+    # https://fasttext.cc/docs/en/supervised-tutorial.html
+    model = fasttext.train_supervised(input_file_path)
+    model.save_model(model_save_path)
+
+
+class Classifier:
+    """分类器"""
+
+    def __init__(self, model_load_path):
+        self.model_load_path = model_load_path
+        self.model = self.load_model()
+
+    def load_model(self):
+        """加载模型"""
+        return fasttext.load_model(self.model_load_path)
+
+    def predict(self, query):
+        """预测 query"""
+
+        # 基于 fasttext api 实现模型预测
+        # https://fasttext.cc/docs/en/supervised-tutorial.html
+        query_intent = self.model.predict(query)
+        # 预测 label 和概率
+        return query_intent[0][0].replace('__label__', ''), query_intent[1][0]
+
+
+if __name__ == '__main__':
+
+    print('开始训练分类器...')
+
+    train_classifier(classifier_corpus_path, classifier_save_path)
+
+    print('分类器训练成功...')  

project5/基于医疗知识图谱的问答系统/student_code/module/graph_matcher.py

+from py2neo import Graph
+from py2neo import Graph
+
+
+class GraphMatcher:
+    """基于 cypher 语句查询数据库"""
+
+    def __init__(self):
+        self.graph = Graph('http://localhost:7474/finance_demo/db/', auth=('neo4j', 'neo4j123'))
+
+    def parse_graph(self, ques_types, entities):
+        """转换成 cypher 语句查询"""
+
+        response = ""
+        for each_ques_type in ques_types:
+            if each_ques_type == 'concept':
+                for entity_name, entity_type in entities.items():
+                    if entity_type == '股票':
+                        cypher_sql = f'MATCH (s:{entity_type})-[r:所属概念]->(c:概念) where s.股票名称 = "{entity_name}" return c.概念名称'
+                        rtn = self.graph.run(cypher_sql).data()
+                        response += f'{entity_name}所属概念是{rtn[0]["c.概念名称"]}' + '\n'
+            elif each_ques_type == 'holder':
+                # 提示：match 股东 - 持有 - 股票
+                for entity_name, entity_type in entities.items():
+                    if entity_type == '股东':
+                        cypher_sql = f'MATCH (s:股东)-[r:持有]->(c:股票) where s.`股东名称` = "{entity_name}" return c.`股票名称`'
+                        rtn = self.graph.run(cypher_sql).data()
+                        response += f'{entity_name}所属持有的股票是{rtn[0]["c.`股票名称`"]}' + '\n'
+                
+            elif each_ques_type == 'industry':
+                # 提示：match 股票 return 行业
+                for entity_name, entity_type in entities.items():
+                    if entity_type == '股票':
+                        cypher_sql = f'MATCH (c:股票) where c.`股票名称` = "{entity_name}" return c.`行业`'
+                        rtn = self.graph.run(cypher_sql).data()
+                        response += f'{entity_name}所属行业是{rtn[0]["c.`行业`"]}' + '\n'
+        return response.strip()
+
+    def predict(self, semantics):
+        """预测 query"""
+        response = self.parse_graph(semantics['ques_types'], semantics['entities'])
+        return response

project5/基于医疗知识图谱的问答系统/student_code/module/semantic_parser.py

+from config import entity_corpus_path, entity_searcher_save_path, contexts
+from config import entity_corpus_path, entity_searcher_save_path, contexts
+import ahocorasick
+import pandas as pd
+import os
+import pickle
+
+
+def build_search_tree(input_folder_path, tree_save_path):
+    """读取股票名称，股东和概念实体，构建 ac 树"""
+    # https://pypi.org/project/pyahocorasick/
+    tree = ahocorasick.Automaton()
+
+    stock_basic = pd.read_csv(os.path.join(input_folder_path, '股票信息.csv'), encoding='gbk')
+    for name in stock_basic["name"]:
+        tree.add_word(name, (name, '股票'))
+    # 遍历 stock_basic，添加 name 即股票名字
+    # 股票名字为key，value表示为具体的实体类型，比如：tree.add_word('股票名A', ('股票名A', '股票'))
+
+    concept = pd.read_csv(os.path.join(input_folder_path, '概念信息.csv'), encoding='gbk')
+    for name in concept["name"]:
+        tree.add_word(name, (name, '概念'))
+    # 遍历 concept，添加 name 即概念名字
+    # 概念名字为key，value表示为具体的实体类型，比如：tree.add_word('概念名A', ('概念名A', '概念'))
+ 
+    holder = pd.read_csv(os.path.join(input_folder_path, '股东信息.csv'), encoding='gbk')
+    for name in holder["股东名称"]:
+        tree.add_word(name, (name, '股东'))
+    # 遍历 holder，添加 股东名称
+    # 股东名称为key，value表示为具体的实体类型，比如：tree.add_word('股东名称A', ('股东名称A', '股东'))
+
+    tree.make_automaton()
+
+    with open(tree_save_path, 'wb') as fout:
+        pickle.dump(tree, fout)
+
+
+class SemanticParser:
+    """实体搜索器"""
+
+    def __init__(self, entity_model_load_path, question_types):
+        self.entity_model_load_path = entity_model_load_path
+        self.entity_model = self.load_model()
+        self.question_types = question_types
+
+    def load_model(self):
+        """加载模型"""
+        with open(self.entity_model_load_path, 'rb') as fin:
+            return pickle.load(fin)
+
+    def predict_question_types(self, query):
+        """判断问题类型，这里只是通过关键词去判断，可以改成分类模型"""
+
+        rtn_ques_types = []
+        for ques_type, kws in self.question_types.items():
+            for each_kw in kws:
+                if each_kw in query:
+                    rtn_ques_types.append(ques_type)
+                    break
+        return rtn_ques_types
+
+    def predict(self, query):
+        """预测 query"""
+
+        rtn = {}
+
+        # 预测类型
+        ques_types = self.predict_question_types(query)
+
+        # 预测实体
+        entities = {}
+        for end_index, (entity_name, entity_type) in self.entity_model.iter(query):
+            entities[entity_name] = entity_type
+
+        if len(ques_types) != 0 and len(entities) != 0:
+            rtn['ques_types'] = ques_types
+            rtn['entities'] = entities
+            # 备份
+            contexts['ques_types'] = ques_types
+            contexts['entities'] = entities
+        elif len(ques_types) != 0:
+            rtn['ques_types'] = ques_types
+            # 备份
+            contexts['ques_types'] = ques_types
+
+            # 从对话历史中继承问题类型
+            rtn['entities'] = contexts['entities']
+        elif len(entities) != 0:
+            # 从对话历史中继承问题类型
+            rtn['ques_types'] = contexts['ques_types']
+
+            rtn['entities'] = entities
+            # 备份
+            contexts['entities'] = entities
+        else:
+            # 如果两个都没有找到，那说明是没有涉及 KG
+            rtn['ques_types'] = []
+            rtn['entities'] = {}
+
+        return rtn
+
+
+if __name__ == '__main__':
+
+    print('开始训练实体搜索树...')
+
+    build_search_tree(entity_corpus_path, entity_searcher_save_path)
+
+    print('实体搜索树训练成功...')
\ No newline at end of file

project5/基于医疗知识图谱的问答系统/student_code/step1_get_data.py

+import tushare as ts
+import tushare as ts
+import pandas as pd
+
+# 和 Tushare 建立连接
+pro = ts.pro_api('fece147e0dfedd3c7cbe10f83a3e848187c8b0430d9def6779681d42')
+
+# 股票基本信息
+# 查询当前所有正常上市交易的股票列表
+# https://waditu.com/document/2?doc_id=25
+# ts_code: TS代码
+# symbol: 股票代码
+# name: 股票名称
+# industry: 行业
+stock_basic = pro.stock_basic(exchange='', list_status='L', fields='ts_code,symbol,name,industry')
+stock_basic.to_csv('./data/knowledge/股票信息.csv', encoding='gbk')
+
+# 概念信息
+# 获取概念股分类
+# https://waditu.com/document/2?doc_id=125
+concept = pro.concept()
+concept.to_csv('./data/knowledge/概念信息.csv', encoding='gbk', index=False)
+
+# 股票概念信息
+# 获取概念下对应的股票
+concept_details = pd.DataFrame(columns=('id', 'concept_name', 'code', 'name'))
+# 在 概念信息.csv 文件中，总共有 358 个 概念
+for i in range(359):
+    concept_id = 'TS' + str(i)
+    # 获取该概念下的全部股票列表
+    # https://waditu.com/document/2?doc_id=126
+    concept_stocks = pro.concept_detail(id=concept_id)
+    concept_details = concept_details.append(concept_stocks)
+concept_details.to_csv('./data/knowledge/股票-概念信息.csv', encoding='gbk')
+
+holder_basic = []
+# 股票持有股东信息
+stock_holders = pd.DataFrame(columns=('ts_code', 'ann_date', 'end_date', 'holder_name', 'hold_amount', 'hold_ratio'))
+# 获取时间段内股票的股东信息
+for each_code in stock_basic['ts_code'].tolist():
+    # 前十大股东：https://waditu.com/document/2?doc_id=61
+    curr_holder = pro.top10_holders(ts_code=each_code, start_date='20180101', end_date='20181231')
+    # 在这里，简单起见，只考虑第一个股东信息
+    stock_holders = stock_holders.append(curr_holder.iloc[0:1])
+    # 加入股东名称
+    # 加入时做清洗，即去除 -，比如将 新华人寿保险股份有限公司-分红-个人分红-018L-FH002深 清洗为 新华人寿保险股份有限公司
+    holder_basic.extend(curr_holder.iloc[0:1]['holder_name'].values.tolist().split('-')[0])
+stock_holders.to_csv('./data/knowledge/股票-股东信息.csv', encoding='gbk')
+
+# 股东信息
+holder_basic_df = pd.DataFrame({
+    '股东名称': list(set(holder_basic))
+})
+holder_basic_df.to_csv('./data/knowledge/股东信息.csv', encoding='gbk', index=False)
+

project5/基于医疗知识图谱的问答系统/student_code/step2_store_to_neo4j.py

+from tqdm import tqdm
+from tqdm import tqdm
+import pandas as pd
+from py2neo import Graph, Node, Relationship, NodeMatcher
+
+# --------------------------- 连接 Neo4j
+# 官方文档：https://py2neo.org/2021.1/
+graph = Graph('http://localhost:7474/finance_demo/db/', auth=('neo4j', 'neo4j123'))
+print(graph)
+
+# --------------------------- 创建实体
+# 股票
+print('创建 股票 实体...')
+stock_basic = pd.read_csv('./data/knowledge/股票信息.csv', encoding='gbk')
+for idx, each_row in tqdm(stock_basic.iterrows()):
+    # 方法说明：https://py2neo.org/2021.1/data/index.html#py2neo.data.Node
+    # 股票 是 label
+    # keyword arguments 是属性，如 TS代码 等
+    each_stock = Node('股票',
+                      TS代码=each_row['ts_code'],
+                      股票代码=each_row['symbol'],
+                      股票名称=each_row['name'],
+                      行业=each_row['industry'])
+    try:
+        # 方法说明：https://py2neo.org/2021.1/workflow.html#py2neo.Transaction.create
+        graph.create(each_stock)
+    except Exception as e:
+        print(f'Error: {e}, data idx: {idx}, data: {each_row}')
+
+# 概念
+print('创建 概念 实体...')
+concept = pd.read_csv('./data/knowledge/概念信息.csv', encoding='gbk')
+for idx, each_row in tqdm(concept.iterrows()):
+    each_concept = Node('概念',
+                        概念代码=each_row['code'],
+                        概念名称=each_row['name'])
+    graph.create(each_concept)
+
+# 股东
+print('创建 股东 实体...')
+holder = pd.read_csv('./data/knowledge/股东信息.csv', encoding='gbk')
+for idx, each_row in tqdm(holder.iterrows()):
+    each_holder = Node('股东',
+                       股东名称=each_row['股东名称'])
+    graph.create(each_holder)
+
+# --------------------------- 创建关系
+# 方法说明：https://py2neo.org/2021.1/matching.html#py2neo.NodeMatcher
+matcher = NodeMatcher(graph)
+
+# 股票-概念
+print('创建 股票-概念 关系...')
+stock_concept = pd.read_csv('./data/knowledge/股票-概念信息.csv', encoding='gbk')
+for idx, each_row in tqdm(stock_concept.iterrows()):
+    node1 = matcher.match('股票', TS代码=each_row['ts_code']).first()
+    node2 = matcher.match('概念', 概念代码=each_row['id']).first()
+    if node1 is not None and node2 is not None:
+        # 方法说明：https://py2neo.org/2021.1/data/index.html#py2neo.data.Relationship
+        # 格式：Relationship(start_node, type, end_node)
+        r = Relationship(node1, '所属概念', node2)
+        graph.create(r)
+
+# 股票-股东
+print('创建 股票-股东 关系...')
+stock_holder = pd.read_csv('./data/knowledge/股票-股东信息.csv', encoding='gbk')
+for idx, each_row in tqdm(stock_holder.iterrows()):
+    # first() 方法返回第一个匹配的 Node，如果找不到则返回 None
+    node1 = matcher.match("股票", TS代码=each_row['ts_code']).first()
+    node2 = matcher.match("股东", 股东名称=each_row['holder_name'].split('-')[0]).first()
+    if node1 is not None and node2 is not None:
+        r = Relationship(node2, '持有', node1,
+                         ann_date=each_row['ann_date'],
+                         end_date=each_row['end_date'],
+                         hold_amount=each_row['hold_amount'],
+                         hold_ratio=each_row['hold_ratio'])
+        graph.create(r)
+
+print('实体 关系 导入成功...')
\ No newline at end of file