skip-gram.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import time\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import random\n",
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('E:/python_work/handson_ml/datasets/Javasplittedwords.txt','r',encoding='utf-8') as f:\n",
    "    text = f.read()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = text.split(' ')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "words_count = Counter(text)\n",
    "words = [w for w in text if words_count[w] > 50]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({'熟练掌握': 14393,\n",
       "         'java': 16940,\n",
       "         '熟悉': 117424,\n",
       "         'python': 7767,\n",
       "         'shell': 3494,\n",
       "         '熟练使用': 20556,\n",
       "         'git': 1728,\n",
       "         'svn': 1210,\n",
       "         '能够': 29684,\n",
       "         '发现': 4603,\n",
       "         '问题': 21185,\n",
       "         '精准': 985,\n",
       "         '定位问题': 195,\n",
       "         '快速': 9775,\n",
       "         '解决问题': 8020,\n",
       "         'jvm': 844,\n",
       "         '优化': 35263,\n",
       "         '经验': 111298,\n",
       "         '分析': 55814,\n",
       "         '找出': 311,\n",
       "         '服务器': 4611,\n",
       "         '代码': 17103,\n",
       "         '存在': 953,\n",
       "         '具有': 53927,\n",
       "         '服务端': 1594,\n",
       "         '架构': 14772,\n",
       "         '理解': 13614,\n",
       "         '掌握': 13518,\n",
       "         '理论': 2763,\n",
       "         '实际操作': 242,\n",
       "         'dubbo': 472,\n",
       "         'thrift': 144,\n",
       "         '工具': 14214,\n",
       "         'mysql': 14202,\n",
       "         'redis': 4592,\n",
       "         '并能': 3691,\n",
       "         '性能': 11192,\n",
       "         'kafka': 876,\n",
       "         'zeromq': 32,\n",
       "         '使用': 19057,\n",
       "         '大数据': 10049,\n",
       "         '高': 9693,\n",
       "         '并发': 4444,\n",
       "         '参与': 24440,\n",
       "         '开源': 6643,\n",
       "         '项目': 49456,\n",
       "         '账号': 756,\n",
       "         '技术': 60222,\n",
       "         '博客': 790,\n",
       "         'zookeeper': 518,\n",
       "         '锁': 81,\n",
       "         '\\n工作职责': 7755,\n",
       "         '负责': 118261,\n",
       "         '应用': 20545,\n",
       "         '规划设计': 810,\n",
       "         '现状': 773,\n",
       "         '梳理': 2240,\n",
       "         '差异': 437,\n",
       "         '迁移': 210,\n",
       "         '方案设计': 1618,\n",
       "         '工作': 116266,\n",
       "         '重大项目': 147,\n",
       "         '评审': 2405,\n",
       "         '跟踪': 10465,\n",
       "         '规划': 19939,\n",
       "         '后端': 4317,\n",
       "         '系统': 35596,\n",
       "         '把握': 5798,\n",
       "         '复杂': 2503,\n",
       "         '设计': 99519,\n",
       "         '确保': 6980,\n",
       "         '质量': 8869,\n",
       "         '现有': 4315,\n",
       "         '解决': 12984,\n",
       "         '中': 19974,\n",
       "         '关键问题': 250,\n",
       "         '技术难题': 1376,\n",
       "         '方向': 7248,\n",
       "         '预研': 693,\n",
       "         '选型': 1968,\n",
       "         '相关': 103998,\n",
       "         '技术规范': 293,\n",
       "         '标准': 7224,\n",
       "         '制定': 20683,\n",
       "         '推广': 14400,\n",
       "         '配合': 13197,\n",
       "         '研发': 15873,\n",
       "         '测试': 22136,\n",
       "         '运维': 3412,\n",
       "         '部门': 15799,\n",
       "         '团队': 65072,\n",
       "         '有序': 267,\n",
       "         '推进': 6436,\n",
       "         '落地': 3814,\n",
       "         '实施': 14304,\n",
       "         '协同': 2197,\n",
       "         '数据': 39647,\n",
       "         '安全': 7400,\n",
       "         '领域': 8427,\n",
       "         '推动': 10013,\n",
       "         '总体': 634,\n",
       "         '任职': 45391,\n",
       "         '要求': 72070,\n",
       "         '全日制': 2883,\n",
       "         '本科': 23277,\n",
       "         '以上学历': 41531,\n",
       "         '计算机': 21583,\n",
       "         '专业': 34273,\n",
       "         '扎实': 7595,\n",
       "         '基础理论': 255,\n",
       "         '知识': 10827,\n",
       "         '工作经验': 36585,\n",
       "         '开发': 113404,\n",
       "         '工作岗位': 108,\n",
       "         '实际': 6834,\n",
       "         '架构设计': 6342,\n",
       "         '集成': 1752,\n",
       "         '交易': 2137,\n",
       "         '服务': 19470,\n",
       "         '项目经验': 3859,\n",
       "         '优先': 83547,\n",
       "         '深入': 6434,\n",
       "         '了解': 35429,\n",
       "         '设计模式': 4368,\n",
       "         '精通': 29541,\n",
       "         'c++': 8199,\n",
       "         '量': 1492,\n",
       "         '高性能': 2322,\n",
       "         '分布式系统': 1743,\n",
       "         '调': 4849,\n",
       "         '优': 4780,\n",
       "         'c': 13785,\n",
       "         '基础': 11198,\n",
       "         '多线程': 4047,\n",
       "         '集合': 485,\n",
       "         '框架': 19936,\n",
       "         '缓存': 3118,\n",
       "         '消息': 1459,\n",
       "         '机制': 3781,\n",
       "         '分布式服务': 127,\n",
       "         '开发框架': 2396,\n",
       "         '建模': 3375,\n",
       "         '认真': 7759,\n",
       "         '积极主动': 5816,\n",
       "         '细致': 4229,\n",
       "         '严谨': 3597,\n",
       "         '善于': 16389,\n",
       "         '思考': 4291,\n",
       "         '总结': 4085,\n",
       "         '良好': 66822,\n",
       "         '合作': 38375,\n",
       "         '沟通': 63734,\n",
       "         '表达能力': 8956,\n",
       "         '系统架构': 3005,\n",
       "         '师资格': 7,\n",
       "         '\\n岗位职责': 38401,\n",
       "         '独立': 25916,\n",
       "         '完成': 49129,\n",
       "         '仓储物流': 62,\n",
       "         '业务流程': 2368,\n",
       "         '开发设计': 460,\n",
       "         '编写': 21144,\n",
       "         '模块': 9134,\n",
       "         '概要': 951,\n",
       "         '说明书': 1300,\n",
       "         '详细设计': 1807,\n",
       "         '数据库': 20934,\n",
       "         '帮助': 3298,\n",
       "         '成员': 2803,\n",
       "         '提高': 6998,\n",
       "         '能力': 152861,\n",
       "         '资格': 15236,\n",
       "         '主流': 8344,\n",
       "         'springmvc': 1138,\n",
       "         'mybatis': 2015,\n",
       "         '微': 1559,\n",
       "         'spring': 4891,\n",
       "         'tomcat': 1963,\n",
       "         '产品': 150293,\n",
       "         '部署': 3678,\n",
       "         '进行': 54195,\n",
       "         '中间件': 1135,\n",
       "         '具备': 47207,\n",
       "         '导出': 64,\n",
       "         '最好': 1850,\n",
       "         'linux': 15292,\n",
       "         '环境': 10573,\n",
       "         '延展性': 25,\n",
       "         '可扩展性': 411,\n",
       "         '流量': 1850,\n",
       "         '访问量': 290,\n",
       "         '容错性': 23,\n",
       "         '负载': 1644,\n",
       "         '公司': 77897,\n",
       "         '软件': 18157,\n",
       "         '商店': 159,\n",
       "         '游戏': 11243,\n",
       "         '中心': 1744,\n",
       "         '论坛': 1046,\n",
       "         '支付': 3905,\n",
       "         '云': 2981,\n",
       "         '开发者': 1153,\n",
       "         '平台': 30953,\n",
       "         '系统分析': 1474,\n",
       "         '实现': 17458,\n",
       "         '持续改善': 706,\n",
       "         '需求分析': 9011,\n",
       "         '持续': 9570,\n",
       "         '完善': 7669,\n",
       "         '过程': 10667,\n",
       "         '研究': 13598,\n",
       "         '实践': 1730,\n",
       "         '原理': 6460,\n",
       "         '语言': 15119,\n",
       "         '编程': 17347,\n",
       "         'struts': 1584,\n",
       "         'hibernate': 2195,\n",
       "         '迭代': 4430,\n",
       "         '模式': 5242,\n",
       "         '自动': 414,\n",
       "         '构建': 3478,\n",
       "         '常用工具': 420,\n",
       "         '以下': 2446,\n",
       "         '级用户': 167,\n",
       "         '海量': 2573,\n",
       "         '互联网': 41861,\n",
       "         'memcached': 954,\n",
       "         'mina': 126,\n",
       "         'jquery': 6362,\n",
       "         'ajax': 4641,\n",
       "         'javascript': 11082,\n",
       "         'html5': 6301,\n",
       "         '前端': 12431,\n",
       "         '用户': 38091,\n",
       "         '系统设计': 2801,\n",
       "         '\\n通用': 7,\n",
       "         '技能': 4954,\n",
       "         '从事': 4376,\n",
       "         '超过': 942,\n",
       "         '金融': 8267,\n",
       "         '发展趋势': 2093,\n",
       "         '学习': 26701,\n",
       "         '创新能力': 1740,\n",
       "         '强': 34815,\n",
       "         '迅速': 1374,\n",
       "         '评估': 5537,\n",
       "         '适用': 183,\n",
       "         '场景': 3333,\n",
       "         '大局观': 257,\n",
       "         '逻辑': 9355,\n",
       "         '碰到问题': 11,\n",
       "         '时有': 19,\n",
       "         '清晰': 8466,\n",
       "         '思路': 4968,\n",
       "         '语言表达': 2878,\n",
       "         '精神': 21267,\n",
       "         '优秀': 18603,\n",
       "         '文档': 24878,\n",
       "         '撰写': 17280,\n",
       "         '软件工程': 1889,\n",
       "         '电子信息': 324,\n",
       "         '数学': 3354,\n",
       "         '统招': 2176,\n",
       "         '专业技能': 413,\n",
       "         '至少': 12925,\n",
       "         '满足': 1934,\n",
       "         '下述': 13,\n",
       "         '金融机构': 588,\n",
       "         '管理': 32818,\n",
       "         '办法': 281,\n",
       "         '信用卡': 361,\n",
       "         '核心': 8644,\n",
       "         '银行卡': 56,\n",
       "         '前置': 23,\n",
       "         '银行': 3900,\n",
       "         '贷款': 534,\n",
       "         '审批': 576,\n",
       "         '主导': 2580,\n",
       "         '规范': 10651,\n",
       "         '清': 226,\n",
       "         '结算': 1713,\n",
       "         '分期': 232,\n",
       "         '个人消费': 26,\n",
       "         '金融产品': 1858,\n",
       "         '大数据处理': 724,\n",
       "         '流式': 114,\n",
       "         '计算': 1747,\n",
       "         '实时': 1528,\n",
       "         '成功': 4812,\n",
       "         '低': 695,\n",
       "         '延时': 21,\n",
       "         '案例': 3182,\n",
       "         '规则': 1658,\n",
       "         '引擎': 2530,\n",
       "         '反': 546,\n",
       "         '欺诈': 697,\n",
       "         'oracle': 5391,\n",
       "         '水平': 1505,\n",
       "         '垂直': 397,\n",
       "         '分库': 138,\n",
       "         'sas': 1005,\n",
       "         '数据分析': 14327,\n",
       "         '深入理解': 2294,\n",
       "         '业务': 36158,\n",
       "         '风险': 4588,\n",
       "         '改进': 9657,\n",
       "         '定位': 3630,\n",
       "         '故障': 1278,\n",
       "         '原因': 505,\n",
       "         '保障': 1971,\n",
       "         '可用': 1266,\n",
       "         '深入分析': 389,\n",
       "         '瓶颈': 595,\n",
       "         '容量': 180,\n",
       "         '手段': 1329,\n",
       "         '提升': 12575,\n",
       "         '整体': 8397,\n",
       "         '职位': 12680,\n",
       "         '客户': 43131,\n",
       "         '意识': 15943,\n",
       "         '强烈': 8288,\n",
       "         '责任心': 18973,\n",
       "         '使命感': 163,\n",
       "         '执行力': 5950,\n",
       "         '富有': 4422,\n",
       "         '拥有': 7208,\n",
       "         '包括': 13800,\n",
       "         '虚拟机': 108,\n",
       "         '网络通信': 483,\n",
       "         '常用': 11558,\n",
       "         '数据传输': 109,\n",
       "         '协议': 5555,\n",
       "         '操作系统': 4253,\n",
       "         '故障分析': 124,\n",
       "         'jetty': 221,\n",
       "         '脚本语言': 1800,\n",
       "         '\\n岗位': 4608,\n",
       "         '后台': 7102,\n",
       "         '架构师': 508,\n",
       "         '研发经理': 141,\n",
       "         '工作地': 73,\n",
       "         '北京': 1907,\n",
       "         '岗位职责': 6305,\n",
       "         '符合': 2345,\n",
       "         '需求': 55003,\n",
       "         '伸缩': 35,\n",
       "         '规模': 820,\n",
       "         '流程': 21864,\n",
       "         '运营': 24883,\n",
       "         '按时': 2995,\n",
       "         '高质量': 3439,\n",
       "         '发布': 4952,\n",
       "         '岗位': 19965,\n",
       "         '基本功': 444,\n",
       "         '服务端开发': 612,\n",
       "         '级别': 558,\n",
       "         '上线': 9021,\n",
       "         '常见': 3676,\n",
       "         '热爱': 9841,\n",
       "         '发展': 16870,\n",
       "         '进取精神': 258,\n",
       "         '创新': 9633,\n",
       "         '主动性': 1607,\n",
       "         '流程管理': 238,\n",
       "         '团队管理': 3311,\n",
       "         '加分': 4085,\n",
       "         '新': 7299,\n",
       "         '特性': 2931,\n",
       "         '分表': 215,\n",
       "         '电商': 6192,\n",
       "         '数据库系统': 580,\n",
       "         'hadoop': 4662,\n",
       "         'spark': 2983,\n",
       "         'storm': 1159,\n",
       "         '数据处理': 1206,\n",
       "         '维护': 27059,\n",
       "         '产品质量': 1118,\n",
       "         '库': 2359,\n",
       "         '数据结构': 4648,\n",
       "         '算法': 9501,\n",
       "         '网络': 15066,\n",
       "         '基础知识': 1567,\n",
       "         '极强': 1809,\n",
       "         '压力': 11751,\n",
       "         '承担': 4469,\n",
       "         '上进心': 1514,\n",
       "         '挑战性': 803,\n",
       "         '充满': 2227,\n",
       "         '激情': 5369,\n",
       "         'php': 10925,\n",
       "         '门': 2020,\n",
       "         '移动': 17429,\n",
       "         '端': 10642,\n",
       "         '产品研发': 2142,\n",
       "         '全': 2033,\n",
       "         '栈': 425,\n",
       "         '对开': 51,\n",
       "         '源': 247,\n",
       "         '兴趣': 3090,\n",
       "         '爱好': 517,\n",
       "         '对应': 898,\n",
       "         '模型': 5098,\n",
       "         '数据模型': 942,\n",
       "         '物理模型': 117,\n",
       "         '带领': 2077,\n",
       "         '产品技术架构': 202,\n",
       "         '路标': 62,\n",
       "         '版本': 2553,\n",
       "         '得到': 1165,\n",
       "         '遵从': 49,\n",
       "         '软件开发': 5345,\n",
       "         '项目管理': 5878,\n",
       "         '运作': 1888,\n",
       "         '体验': 6043,\n",
       "         '门户网站': 302,\n",
       "         '社区': 2544,\n",
       "         '网站': 17769,\n",
       "         '电子商务': 2935,\n",
       "         '安全性': 585,\n",
       "         '制订': 1358,\n",
       "         '解决方案': 8375,\n",
       "         '指导': 6514,\n",
       "         '技术指导': 440,\n",
       "         '集群': 1643,\n",
       "         '配置': 3664,\n",
       "         '复制': 163,\n",
       "         '配置管理': 418,\n",
       "         '运行': 2645,\n",
       "         '较强': 37265,\n",
       "         '最新': 2601,\n",
       "         '发展趋向': 17,\n",
       "         '经验丰富': 498,\n",
       "         '重构': 1525,\n",
       "         '业务规划': 415,\n",
       "         '方案': 22489,\n",
       "         '攻关': 1851,\n",
       "         '请求': 319,\n",
       "         '数': 1446,\n",
       "         '处理': 13507,\n",
       "         '功能': 15285,\n",
       "         '模板': 895,\n",
       "         '公用': 134,\n",
       "         '收集': 9721,\n",
       "         '整理': 8193,\n",
       "         '资料': 5383,\n",
       "         '技术成果': 25,\n",
       "         '审核': 4070,\n",
       "         '把关': 284,\n",
       "         '专科': 2859,\n",
       "         '软件架构': 431,\n",
       "         '体系结构': 439,\n",
       "         '思想': 2641,\n",
       "         '全面': 3495,\n",
       "         '知识结构': 154,\n",
       "         '网络安全': 1316,\n",
       "         '任务': 10115,\n",
       "         '进度': 6238,\n",
       "         '控': 5397,\n",
       "         '对外': 2520,\n",
       "         '协作': 13632,\n",
       "         '销售': 26037,\n",
       "         '识别': 1940,\n",
       "         '商业机会': 153,\n",
       "         '提供': 25624,\n",
       "         '售前': 2737,\n",
       "         '技术支持': 2272,\n",
       "         '成为': 2432,\n",
       "         '桥梁': 171,\n",
       "         '程度': 541,\n",
       "         '加强': 454,\n",
       "         '客户关系': 2238,\n",
       "         '实施方案': 500,\n",
       "         '书': 471,\n",
       "         '投标书': 131,\n",
       "         '技术文案': 32,\n",
       "         '行业': 30182,\n",
       "         '专业本科': 5259,\n",
       "         '经历': 2857,\n",
       "         '细分': 461,\n",
       "         '业务知识': 223,\n",
       "         '调研': 7664,\n",
       "         '项目经理': 1287,\n",
       "         '人数': 296,\n",
       "         '选择': 1569,\n",
       "         '影响': 870,\n",
       "         '部门总经理': 14,\n",
       "         '决策': 2684,\n",
       "         '产生': 607,\n",
       "         '云计算': 1819,\n",
       "         '文字': 8213,\n",
       "         '图示': 44,\n",
       "         '清楚': 455,\n",
       "         '表达': 4166,\n",
       "         '意图': 317,\n",
       "         '熟练': 8653,\n",
       "         '娴熟': 386,\n",
       "         '技巧': 5352,\n",
       "         '敬业精神': 2328,\n",
       "         '认知': 1219,\n",
       "         '热情': 5438,\n",
       "         '集团': 1377,\n",
       "         '资金': 1500,\n",
       "         '负责项目': 2439,\n",
       "         '汇总': 1547,\n",
       "         '生命周期': 2414,\n",
       "         '清算': 214,\n",
       "         '直连': 15,\n",
       "         '管理系统': 963,\n",
       "         '建设': 7321,\n",
       "         '日语': 208,\n",
       "         '方法': 7845,\n",
       "         '风险管理': 899,\n",
       "         '项目计划': 612,\n",
       "         '校准': 11,\n",
       "         '汇报': 1256,\n",
       "         '丰富': 9908,\n",
       "         '表达清晰': 120,\n",
       "         '协调': 25273,\n",
       "         '\\n职位': 10034,\n",
       "         '描述': 14206,\n",
       "         '深刻理解': 2713,\n",
       "         '同事': 2099,\n",
       "         '紧密配合': 969,\n",
       "         '对系统': 801,\n",
       "         '子系统': 321,\n",
       "         '做出': 1246,\n",
       "         '前瞻性': 1019,\n",
       "         '电商运营': 158,\n",
       "         '关键': 1616,\n",
       "         '可靠性': 566,\n",
       "         '可用性': 2740,\n",
       "         '可维护性': 341,\n",
       "         '深度': 3353,\n",
       "         '简化': 51,\n",
       "         '积极参与': 286,\n",
       "         '人员': 7865,\n",
       "         '招聘': 4015,\n",
       "         '建议': 6381,\n",
       "         '加速': 150,\n",
       "         '成长': 4751,\n",
       "         '大学本科': 3096,\n",
       "         '大客户': 692,\n",
       "         '互联': 281,\n",
       "         '信息': 13418,\n",
       "         '反应': 801,\n",
       "         '敏捷': 4398,\n",
       "         '面向对象': 3182,\n",
       "         '佳': 3621,\n",
       "         '设计工具': 1389,\n",
       "         '深入研究': 791,\n",
       "         '熟知': 695,\n",
       "         '不同': 3970,\n",
       "         '流行': 1790,\n",
       "         '优缺点': 302,\n",
       "         'visio': 2328,\n",
       "         '高度': 4296,\n",
       "         '领袖': 97,\n",
       "         '必须': 1892,\n",
       "         '系统软件': 232,\n",
       "         '商业运营': 40,\n",
       "         '难点': 1024,\n",
       "         'nosql': 2380,\n",
       "         '实战经验': 742,\n",
       "         '协助': 16146,\n",
       "         '组织': 14377,\n",
       "         '数据管理': 214,\n",
       "         '范围': 1013,\n",
       "         '项目成果': 50,\n",
       "         '企业级': 684,\n",
       "         '更新': 4172,\n",
       "         '企业': 10828,\n",
       "         '充分': 2052,\n",
       "         '支撑': 1523,\n",
       "         '兼顾': 140,\n",
       "         '扩展': 978,\n",
       "         '关联': 466,\n",
       "         '衔接': 318,\n",
       "         '未来': 2048,\n",
       "         '编制': 3570,\n",
       "         '战略': 4296,\n",
       "         '中长期': 227,\n",
       "         '建设项目': 60,\n",
       "         '制度': 3253,\n",
       "         '标准规范': 156,\n",
       "         '运用': 2557,\n",
       "         '贯彻执行': 212,\n",
       "         '层面': 669,\n",
       "         '检查': 1284,\n",
       "         '督导': 246,\n",
       "         '可行性研究': 123,\n",
       "         '总体方案': 125,\n",
       "         '关键技术': 732,\n",
       "         '验证': 1787,\n",
       "         '治理': 204,\n",
       "         '体系': 8212,\n",
       "         '元数据': 215,\n",
       "         '主': 618,\n",
       "         '含': 1210,\n",
       "         '完整': 3572,\n",
       "         '交付': 2475,\n",
       "         '骨干': 209,\n",
       "         'hbase': 1927,\n",
       "         '较大': 3060,\n",
       "         '自我管理': 512,\n",
       "         '口头': 1675,\n",
       "         'data': 685,\n",
       "         '利用': 2940,\n",
       "         'css': 9152,\n",
       "         '界面': 5187,\n",
       "         '负责制': 27,\n",
       "         'h5': 1908,\n",
       "         '部分': 2470,\n",
       "         '增加': 1136,\n",
       "         '动态': 3587,\n",
       "         '讨论': 2031,\n",
       "         'web前端': 3184,\n",
       "         '给出': 1779,\n",
       "         '前台': 1555,\n",
       "         '领导': 5874,\n",
       "         '交办': 3205,\n",
       "         'css3': 3003,\n",
       "         '开发技术': 2729,\n",
       "         '效果图': 302,\n",
       "         '页面': 8784,\n",
       "         '指定': 754,\n",
       "         '动态效果': 90,\n",
       "         'js': 7196,\n",
       "         '面向对象编程': 1266,\n",
       "         '兼容': 1000,\n",
       "         '插件': 643,\n",
       "         '浏览器': 4034,\n",
       "         '兼容性问题': 591,\n",
       "         '手机': 3603,\n",
       "         '版': 382,\n",
       "         '屏幕': 77,\n",
       "         '分辨率': 84,\n",
       "         '适配': 422,\n",
       "         '态度': 3510,\n",
       "         '控制器': 79,\n",
       "         '搭建': 5137,\n",
       "         '稳定性': 1637,\n",
       "         '组网': 52,\n",
       "         '遇到': 1376,\n",
       "         '网络协议': 952,\n",
       "         '应用层': 193,\n",
       "         '已有': 1017,\n",
       "         '基础架构': 427,\n",
       "         '升级': 2171,\n",
       "         '不断': 6357,\n",
       "         '效率': 3241,\n",
       "         '底层': 1544,\n",
       "         '支持': 6666,\n",
       "         '保证': 7008,\n",
       "         '数据产品': 1965,\n",
       "         '数据挖掘': 4056,\n",
       "         '机器学习': 2350,\n",
       "         '系统性': 408,\n",
       "         '培训': 14001,\n",
       "         '开发人员': 2273,\n",
       "         '出现': 1312,\n",
       "         '不限': 4825,\n",
       "         '化': 2520,\n",
       "         '轻量化': 16,\n",
       "         '方式': 4205,\n",
       "         '方面': 6046,\n",
       "         '路线': 1059,\n",
       "         '任责': 3,\n",
       "         '学历': 3981,\n",
       "         '大学': 805,\n",
       "         '更好': 1247,\n",
       "         '性格': 1778,\n",
       "         '积极': 7028,\n",
       "         '乐观': 2238,\n",
       "         '诚信': 796,\n",
       "         '驱动': 3868,\n",
       "         '进取心': 1560,\n",
       "         '求知欲': 802,\n",
       "         '计划': 11631,\n",
       "         '带': 1178,\n",
       "         '背景': 6063,\n",
       "         '人士': 266,\n",
       "         '\\n负责': 2985,\n",
       "         '前端架构': 484,\n",
       "         '高级': 1073,\n",
       "         '通用': 822,\n",
       "         '组件': 2432,\n",
       "         '操作': 7542,\n",
       "         '站点': 332,\n",
       "         '模块化': 740,\n",
       "         '组件化': 185,\n",
       "         'mvc': 2939,\n",
       "         'angularjs': 1438,\n",
       "         '详细': 1378,\n",
       "         '任意': 731,\n",
       "         '自动化': 6421,\n",
       "         '阅读': 3498,\n",
       "         '源码': 785,\n",
       "         '愿意': 2691,\n",
       "         '一起': 4937,\n",
       "         '大型项目': 934,\n",
       "         '发掘': 1202,\n",
       "         '主持': 639,\n",
       "         '原型': 9864,\n",
       "         '统一': 948,\n",
       "         '工程师': 2149,\n",
       "         '计算机软件': 923,\n",
       "         '电子商务系统': 44,\n",
       "         '企业应用': 193,\n",
       "         '面向对象分析': 373,\n",
       "         '过程控制': 55,\n",
       "         '高可用性': 402,\n",
       "         '数据源': 233,\n",
       "         '访问': 1212,\n",
       "         '实时性': 24,\n",
       "         '内容': 10701,\n",
       "         '过滤': 156,\n",
       "         '作弊': 94,\n",
       "         '检测': 918,\n",
       "         '监控': 5803,\n",
       "         '存储': 3797,\n",
       "         '非常': 1615,\n",
       "         '敏感性': 1215,\n",
       "         '预测': 756,\n",
       "         '技术创新': 219,\n",
       "         '专注': 1590,\n",
       "         '精益求精': 304,\n",
       "         '淘宝': 1381,\n",
       "         '猫': 920,\n",
       "         '团': 795,\n",
       "         '京东': 785,\n",
       "         '介绍': 1818,\n",
       "         '拜': 129,\n",
       "         '单车': 32,\n",
       "         '一家': 1230,\n",
       "         '发': 503,\n",
       "         '市场': 24076,\n",
       "         '起着': 5,\n",
       "         '重要': 2283,\n",
       "         '作用': 190,\n",
       "         '希望': 3298,\n",
       "         '产品线': 2654,\n",
       "         '范围内': 281,\n",
       "         '发挥': 527,\n",
       "         '价值': 3270,\n",
       "         '刚刚': 50,\n",
       "         '等待': 92,\n",
       "         '需要': 7653,\n",
       "         '统计分析': 1368,\n",
       "         '极大': 291,\n",
       "         '加入': 4377,\n",
       "         '功底': 6958,\n",
       "         '严苛': 14,\n",
       "         '商业智能': 186,\n",
       "         '浓厚兴趣': 874,\n",
       "         '职责': 2235,\n",
       "         '核心技术': 819,\n",
       "         '决策树': 200,\n",
       "         '向量': 19,\n",
       "         '机': 392,\n",
       "         '线性': 58,\n",
       "         '回归': 493,\n",
       "         '谛': 2,\n",
       "         '神经网络': 140,\n",
       "         '实时处理': 81,\n",
       "         '顺利进行': 353,\n",
       "         'scala': 662,\n",
       "         '编程语言': 2033,\n",
       "         '容器': 400,\n",
       "         '逻辑思维': 8262,\n",
       "         '敏感': 3988,\n",
       "         '新技术': 4251,\n",
       "         '感兴趣': 1152,\n",
       "         '跟进': 7857,\n",
       "         '商业': 3951,\n",
       "         '国内外': 1664,\n",
       "         '前沿技术': 909,\n",
       "         '形成': 2515,\n",
       "         '私有云': 149,\n",
       "         '混合': 279,\n",
       "         '里': 674,\n",
       "         '初': 137,\n",
       "         '高级开发工程师': 41,\n",
       "         '贡献': 561,\n",
       "         '虚拟化': 572,\n",
       "         '数理': 72,\n",
       "         '思维': 10543,\n",
       "         '条理清晰': 374,\n",
       "         '精确定位': 1,\n",
       "         '踏实': 4054,\n",
       "         '科技': 2052,\n",
       "         '英语': 2891,\n",
       "         '单元测试': 2077,\n",
       "         '外部': 1754,\n",
       "         '内部': 6297,\n",
       "         '严格': 851,\n",
       "         '合理': 3770,\n",
       "         '分工': 204,\n",
       "         '引进': 321,\n",
       "         '培养': 2048,\n",
       "         '本部门': 354,\n",
       "         '绩效': 1752,\n",
       "         '考评': 173,\n",
       "         '管理工作': 2553,\n",
       "         '考虑': 14083,\n",
       "         '脚本': 3587,\n",
       "         '分布': 246,\n",
       "         '应用逻辑': 37,\n",
       "         '负责管理': 412,\n",
       "         '左右': 993,\n",
       "         '相应': 4610,\n",
       "         '演进': 372,\n",
       "         '领域建模': 40,\n",
       "         '理念': 2413,\n",
       "         '均衡': 1087,\n",
       "         '队列': 537,\n",
       "         '搜索引擎': 936,\n",
       "         '最前沿': 189,\n",
       "         '高德': 95,\n",
       "         '地图搜索': 4,\n",
       "         '信息检索': 116,\n",
       "         '修改': 1825,\n",
       "         '相关性': 84,\n",
       "         '策略': 9693,\n",
       "         '灵活': 1630,\n",
       "         '易': 544,\n",
       "         '检索系统': 32,\n",
       "         '习惯': 6065,\n",
       "         '深刻': 2681,\n",
       "         '实践经验': 1707,\n",
       "         '充满热情': 737,\n",
       "         '新事物': 789,\n",
       "         '转换': 374,\n",
       "         '协助项目经理': 430,\n",
       "         '控制': 4791,\n",
       "         '辅助': 1188,\n",
       "         '开展': 3574,\n",
       "         '员工': 15286,\n",
       "         '产品开发': 5405,\n",
       "         '技术培训': 402,\n",
       "         '大专': 9568,\n",
       "         '软件产品': 2129,\n",
       "         '总体规划': 266,\n",
       "         '队伍': 307,\n",
       "         '技术标准': 247,\n",
       "         '很好': 2470,\n",
       "         '当前': 557,\n",
       "         '抗压': 6901,\n",
       "         '困难': 341,\n",
       "         '带领团队': 1638,\n",
       "         '担任': 512,\n",
       "         '数据中心': 626,\n",
       "         '整体规划': 677,\n",
       "         '设计管理': 90,\n",
       "         '施工': 208,\n",
       "         '机房': 257,\n",
       "         '综合': 2417,\n",
       "         '布线': 257,\n",
       "         '弱电': 74,\n",
       "         '直流电源': 6,\n",
       "         '油机': 4,\n",
       "         '供配电': 9,\n",
       "         '暖通': 29,\n",
       "         '空调': 150,\n",
       "         '给排水': 8,\n",
       "         '布局': 2156,\n",
       "         '装修': 464,\n",
       "         '初设': 9,\n",
       "         '参与项目': 1420,\n",
       "         '设计方案': 1925,\n",
       "         '确定': 1865,\n",
       "         '验收': 1421,\n",
       "         '电气工程及其自动化': 8,\n",
       "         '电力': 197,\n",
       "         '电气': 290,\n",
       "         '国家': 4439,\n",
       "         '设计规范': 1079,\n",
       "         '前期': 1561,\n",
       "         '认识': 2992,\n",
       "         '设备': 3579,\n",
       "         '基本': 6543,\n",
       "         '网络工程师': 45,\n",
       "         '规划系统': 33,\n",
       "         '目标': 12542,\n",
       "         '分析师': 163,\n",
       "         '开放': 1519,\n",
       "         '紧密': 1330,\n",
       "         '转化': 1767,\n",
       "         '风险运营': 12,\n",
       "         '流失': 171,\n",
       "         '贷': 667,\n",
       "         '跨': 2249,\n",
       "         '如期': 221,\n",
       "         '学校': 551,\n",
       "         '产品设计': 11575,\n",
       "         '独产': 69,\n",
       "         '敏感度': 2198,\n",
       "         '视野': 1093,\n",
       "         '敏锐': 5772,\n",
       "         '捕获': 137,\n",
       "         '机会': 6371,\n",
       "         '基本概念': 188,\n",
       "         '决策问题': 46,\n",
       "         '上有': 296,\n",
       "         'rabbitmq': 344,\n",
       "         '信用': 384,\n",
       "         '授权': 291,\n",
       "         '伪冒': 5,\n",
       "         '数据系统': 195,\n",
       "         '定义': 2227,\n",
       "         '扁平化': 386,\n",
       "         '改造': 378,\n",
       "         '资源池': 8,\n",
       "         '国有企业': 9,\n",
       "         '骨干网': 12,\n",
       "         '路由': 525,\n",
       "         '基本原理': 459,\n",
       "         '思科': 159,\n",
       "         '华为': 648,\n",
       "         '交换机': 529,\n",
       "         '\\n精通': 274,\n",
       "         '高可靠性': 191,\n",
       "         '爬虫': 719,\n",
       "         '做': 6773,\n",
       "         '强制': 42,\n",
       "         '地图': 604,\n",
       "         '开放平台': 557,\n",
       "         '查询': 1308,\n",
       "         '鉴': 17,\n",
       "         '权': 48,\n",
       "         '频度': 11,\n",
       "         '额度': 73,\n",
       "         '控制系统': 175,\n",
       "         '防护系统': 8,\n",
       "         '批量': 73,\n",
       "         '索引': 600,\n",
       "         '上下文': 13,\n",
       "         '检索': 252,\n",
       "         '在线': 1996,\n",
       "         '熟': 129,\n",
       "         '系': 342,\n",
       "         '简洁': 275,\n",
       "         '风格': 4448,\n",
       "         '\\n大数据': 32,\n",
       "         '平台架构师': 18,\n",
       "         '责任': 1263,\n",
       "         '产品经理': 4082,\n",
       "         '驾驶舱': 2,\n",
       "         '数据仓库': 2197,\n",
       "         '带过': 102,\n",
       "         '可视化': 1165,\n",
       "         '全局观': 99,\n",
       "         '交流': 3390,\n",
       "         '动手': 812,\n",
       "         '调用': 430,\n",
       "         '会': 7093,\n",
       "         '网络工具': 68,\n",
       "         '分布式文件系统': 83,\n",
       "         'hdfs': 424,\n",
       "         '读写能力': 584,\n",
       "         '提交': 1881,\n",
       "         '补丁': 42,\n",
       "         '成熟': 1619,\n",
       "         '很强': 4761,\n",
       "         '\\n审查': 3,\n",
       "         '构件': 57,\n",
       "         '接口': 4789,\n",
       "         '关系': 6058,\n",
       "         '特别是在': 12,\n",
       "         '测试方法': 377,\n",
       "         '线路': 312,\n",
       "         '质量标准': 87,\n",
       "         '系统测试': 378,\n",
       "         '较深': 1253,\n",
       "         '做好': 5257,\n",
       "         '缺陷': 1004,\n",
       "         '隔离': 51,\n",
       "         '作出': 378,\n",
       "         '提出': 10598,\n",
       "         '改善': 2000,\n",
       "         '测试工具': 1022,\n",
       "         '设计系': 10,\n",
       "         '统': 170,\n",
       "         '扩展性': 441,\n",
       "         '聪明': 368,\n",
       "         '创业': 3679,\n",
       "         '喜欢': 4314,\n",
       "         '挑战': 4899,\n",
       "         '知名': 1930,\n",
       "         '编码': 5755,\n",
       "         '阿里云': 312,\n",
       "         '输出': 3423,\n",
       "         '存储系统': 292,\n",
       "         '拆分': 159,\n",
       "         'elasticsearch': 357,\n",
       "         'springboot': 43,\n",
       "         '归纳': 1478,\n",
       "         '抽象': 1341,\n",
       "         'app': 9369,\n",
       "         '系统核心': 245,\n",
       "         '疑难问题': 584,\n",
       "         ...})"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words_count"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "vocab = set(words)\n",
    "vocab_to_int = {w:c for c ,w in enumerate(vocab)}\n",
    "int_to_vocab = {c: w for c,w in enumerate(vocab)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3038"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vocab_to_int['word']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total words：8614953\n",
      "unique words: 6846\n"
     ]
    }
   ],
   "source": [
    "print('total words：{}'.format(len(words)))\n",
    "print('unique words: {}'.format(len(vocab)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "int_words = [vocab_to_int[w] for w in words]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[6444, 4257, 4506, 477, 580, 244, 4279, 2437, 5083, 983]"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "int_words[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'催收'"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "int_to_vocab[1518]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t = 1e-5\n",
    "threshold = 0.8\n",
    "\n",
    "int_word_counts = Counter(int_words)\n",
    "total_count = len(int_words)\n",
    "# 计算单词频率\n",
    "word_freqs = {w: c/total_count for w, c in int_word_counts.items()}\n",
    "# 计算被删除的概率\n",
    "prob_drop = {w: 1 - np.sqrt(t / word_freqs[w]) for w in int_word_counts}\n",
    "# 对单词进行采样\n",
    "train_words = [w for w in int_words if prob_drop[w] < threshold]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2022807"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_targets(words, idx, window_size=5):\n",
    "    '''\n",
    "    获得input word的上下文单词列表\n",
    "    \n",
    "    参数\n",
    "    ---\n",
    "    words: 单词列表\n",
    "    idx: input word的索引号\n",
    "    window_size: 窗口大小\n",
    "    '''\n",
    "    target_window = np.random.randint(1, window_size+1)\n",
    "    # 这里要考虑input word前面单词不够的情况\n",
    "    start_point = idx - target_window if (idx - target_window) > 0 else 0\n",
    "    end_point = idx + target_window\n",
    "    # output words(即窗口中的上下文单词)\n",
    "    targets = set(words[start_point: idx] + words[idx+1: end_point+1])\n",
    "    return list(targets)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_batches(words, batch_size, window_size=5):\n",
    "    '''\n",
    "    构造一个获取batch的生成器\n",
    "    '''\n",
    "    n_batches = len(words) // batch_size\n",
    "    \n",
    "    # 仅取full batches\n",
    "    words = words[:n_batches*batch_size]\n",
    "    \n",
    "    for idx in range(0, len(words), batch_size):\n",
    "        x, y = [], []\n",
    "        batch = words[idx: idx+batch_size]\n",
    "        for i in range(len(batch)):\n",
    "            batch_x = batch[i]\n",
    "            batch_y = get_targets(batch, i, window_size)\n",
    "            # 由于一个input word会对应多个output word，因此需要长度统一\n",
    "            x.extend([batch_x]*len(batch_y))\n",
    "            y.extend(batch_y)\n",
    "        yield x, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "train_graph = tf.Graph()\n",
    "with train_graph.as_default():\n",
    "    inputs = tf.placeholder(tf.int32, shape=[None], name='inputs')\n",
    "    labels = tf.placeholder(tf.int32, shape=[None, None], name='labels')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "vocab_size = len(int_to_vocab)\n",
    "embedding_size = 200 # 嵌入维度"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6846"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vocab_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with train_graph.as_default():\n",
    "    embedding = tf.Variable(tf.random_uniform([vocab_size,embedding_size],-1,1))\n",
    "    embed = tf.nn.embedding_lookup(embedding,inputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_sampled = 100\n",
    "with train_graph.as_default():\n",
    "    softmax_w = tf.Variable(tf.truncated_normal([vocab_size,embedding_size],stddev=0.1))\n",
    "    softmax_b = tf.Variable(tf.zeros(vocab_size))\n",
    "    \n",
    "    loss = tf.nn.sampled_softmax_loss(softmax_w,softmax_b,labels,embed,n_sampled,vocab_size)\n",
    "    \n",
    "    cost = tf.reduce_mean(loss)\n",
    "    optimizer = tf.train.AdamOptimizer().minimize(cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "with train_graph.as_default():\n",
    "    # 随机挑选一些单词\n",
    "    ## From Thushan Ganegedara's implementation\n",
    "    valid_size = 7 # Random set of words to evaluate similarity on.\n",
    "    valid_window = 100\n",
    "    # pick 8 samples from (0,100) and (1000,1100) each ranges. lower id implies more frequent \n",
    "    valid_examples = np.array(random.sample(range(valid_window), valid_size//2))\n",
    "    valid_examples = np.append(valid_examples, \n",
    "                               random.sample(range(1000,1000+valid_window), valid_size//2))\n",
    "    valid_examples = [\n",
    "                      vocab_to_int['熟悉'],\n",
    "                      vocab_to_int['能力'], \n",
    "                      vocab_to_int['逻辑思维'],\n",
    "                      vocab_to_int['了解']]\n",
    "    \n",
    "    valid_size = len(valid_examples)\n",
    "    # 验证单词集\n",
    "    valid_dataset = tf.constant(valid_examples, dtype=tf.int32)\n",
    "    \n",
    "    # 计算每个词向量的模并进行单位化\n",
    "    norm = tf.sqrt(tf.reduce_sum(tf.square(embedding), 1, keep_dims=True))\n",
    "    normalized_embedding = embedding / norm\n",
    "    # 查找验证单词的词向量\n",
    "    valid_embedding = tf.nn.embedding_lookup(normalized_embedding, valid_dataset)\n",
    "    # 计算余弦相似度\n",
    "    similarity = tf.matmul(valid_embedding, tf.transpose(normalized_embedding))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10 Iteration: 100 Avg. Training loss: 3.9663 0.3036 sec/batch\n",
      "Epoch 1/10 Iteration: 200 Avg. Training loss: 3.6956 0.2889 sec/batch\n",
      "Epoch 1/10 Iteration: 300 Avg. Training loss: 3.8841 0.2871 sec/batch\n",
      "Epoch 1/10 Iteration: 400 Avg. Training loss: 3.7072 0.3501 sec/batch\n",
      "Epoch 1/10 Iteration: 500 Avg. Training loss: 3.5868 0.3471 sec/batch\n",
      "Epoch 1/10 Iteration: 600 Avg. Training loss: 3.6430 0.3575 sec/batch\n",
      "Epoch 1/10 Iteration: 700 Avg. Training loss: 3.4255 0.3616 sec/batch\n",
      "Epoch 1/10 Iteration: 800 Avg. Training loss: 3.4317 0.3543 sec/batch\n",
      "Epoch 1/10 Iteration: 900 Avg. Training loss: 3.4153 0.3720 sec/batch\n",
      "Epoch 1/10 Iteration: 1000 Avg. Training loss: 3.3137 0.3751 sec/batch\n",
      "Nearest to [熟悉]: 优势, 数据源, 集群, 项目管理, 新领域, 定位问题, 须, 小额贷款,\n",
      "Nearest to [能力]: 组件化, 通报, 电子相关, 带领团队, 名称, 广告主, 工程师, 之内,\n",
      "Nearest to [逻辑思维]: 故事, \n",
      "正在, 结构设计, 充分考虑, 管理工具, 提高, 受众, 好奇,\n",
      "Nearest to [了解]: 范, 严格要求, 塑造, 四大, 教育, 出色, 副总经理, 市场策略,\n",
      "Epoch 1/10 Iteration: 1100 Avg. Training loss: 3.2686 0.3695 sec/batch\n",
      "Epoch 1/10 Iteration: 1200 Avg. Training loss: 3.3123 0.3962 sec/batch\n",
      "Epoch 1/10 Iteration: 1300 Avg. Training loss: 3.1527 0.3781 sec/batch\n",
      "Epoch 1/10 Iteration: 1400 Avg. Training loss: 3.0747 0.3905 sec/batch\n",
      "Epoch 1/10 Iteration: 1500 Avg. Training loss: 3.2038 0.3813 sec/batch\n",
      "Epoch 1/10 Iteration: 1600 Avg. Training loss: 3.2911 0.3975 sec/batch\n",
      "Epoch 1/10 Iteration: 1700 Avg. Training loss: 2.9745 0.3746 sec/batch\n",
      "Epoch 1/10 Iteration: 1800 Avg. Training loss: 3.1342 0.3898 sec/batch\n",
      "Epoch 1/10 Iteration: 1900 Avg. Training loss: 2.9352 0.3905 sec/batch\n",
      "Epoch 1/10 Iteration: 2000 Avg. Training loss: 3.0459 0.4947 sec/batch\n",
      "Nearest to [熟悉]: 优势, 数据源, 集群, 项目管理, 新领域, 位置, 定位问题, 须,\n",
      "Nearest to [能力]: 组件化, 通报, 带领团队, 电子相关, 名称, 工程师, 之内, ssl,\n",
      "Nearest to [逻辑思维]: 故事, \n",
      "正在, 结构设计, 充分考虑, 提供商, 管理工具, 提高, 碰撞,\n",
      "Nearest to [了解]: 范, 塑造, 关爱, 严格要求, 四大, 教育, 出色, 原则,\n",
      "Epoch 2/10 Iteration: 2100 Avg. Training loss: 2.7817 0.3056 sec/batch\n",
      "Epoch 2/10 Iteration: 2200 Avg. Training loss: 2.6702 0.4074 sec/batch\n",
      "Epoch 2/10 Iteration: 2300 Avg. Training loss: 2.9330 0.3984 sec/batch\n",
      "Epoch 2/10 Iteration: 2400 Avg. Training loss: 2.9727 0.4999 sec/batch\n",
      "Epoch 2/10 Iteration: 2500 Avg. Training loss: 2.8243 0.4393 sec/batch\n",
      "Epoch 2/10 Iteration: 2600 Avg. Training loss: 3.0719 0.4061 sec/batch\n",
      "Epoch 2/10 Iteration: 2700 Avg. Training loss: 2.8438 0.3921 sec/batch\n",
      "Epoch 2/10 Iteration: 2800 Avg. Training loss: 2.8957 0.4046 sec/batch\n",
      "Epoch 2/10 Iteration: 2900 Avg. Training loss: 2.8859 0.4026 sec/batch\n",
      "Epoch 2/10 Iteration: 3000 Avg. Training loss: 2.7703 0.3922 sec/batch\n",
      "Nearest to [熟悉]: 优势, 集群, 数据源, 位置, 项目管理, 新领域, 定位问题, 服务化,\n",
      "Nearest to [能力]: 组件化, 通报, 电子相关, 带领团队, 名称, 工程师, ssl, 之内,\n",
      "Nearest to [逻辑思维]: 故事, \n",
      "正在, 结构设计, 充分考虑, 提供商, 提高, 碰撞, 管理工具,\n",
      "Nearest to [了解]: 范, 关爱, 塑造, 四大, 严格要求, 教育, 出色, 原则,\n",
      "Epoch 2/10 Iteration: 3100 Avg. Training loss: 2.7395 0.4170 sec/batch\n",
      "Epoch 2/10 Iteration: 3200 Avg. Training loss: 2.9653 0.4002 sec/batch\n",
      "Epoch 2/10 Iteration: 3300 Avg. Training loss: 2.8811 0.4024 sec/batch\n",
      "Epoch 2/10 Iteration: 3400 Avg. Training loss: 2.8006 0.4093 sec/batch\n",
      "Epoch 2/10 Iteration: 3500 Avg. Training loss: 2.7227 0.3990 sec/batch\n",
      "Epoch 2/10 Iteration: 3600 Avg. Training loss: 2.8463 0.4161 sec/batch\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-60-fc7f98b56502>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     15\u001b[0m         \u001b[0mstart\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m         \u001b[1;31m#\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m         \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mbatches\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     18\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m             feed = {inputs: x,\n",
      "\u001b[0;32m<ipython-input-21-cee3a34ea2a0>\u001b[0m in \u001b[0;36mget_batches\u001b[0;34m(words, batch_size, window_size)\u001b[0m\n\u001b[1;32m     13\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m             \u001b[0mbatch_x\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mbatch\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mi\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m---> 15\u001b[0;31m             \u001b[0mbatch_y\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mget_targets\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mwindow_size\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     16\u001b[0m             \u001b[1;31m# 由于一个input word会对应多个output word，因此需要长度统一\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m             \u001b[0mx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mextend\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mbatch_x\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mbatch_y\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-20-360cf480e4f9>\u001b[0m in \u001b[0;36mget_targets\u001b[0;34m(words, idx, window_size)\u001b[0m\n\u001b[1;32m     15\u001b[0m     \u001b[1;31m# output words(即窗口中的上下文单词)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m     \u001b[0mtargets\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mset\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mwords\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mstart_point\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0midx\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m+\u001b[0m \u001b[0mwords\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0midx\u001b[0m\u001b[1;33m+\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m:\u001b[0m \u001b[0mend_point\u001b[0m\u001b[1;33m+\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m     \u001b[1;32mreturn\u001b[0m \u001b[0mlist\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mtargets\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "epochs = 10 # 迭代轮数\n",
    "batch_size = 1000 # batch大小\n",
    "window_size = 10 # 窗口大小\n",
    "\n",
    "with train_graph.as_default():\n",
    "    saver = tf.train.Saver() # 文件存储\n",
    "\n",
    "with tf.Session(graph=train_graph) as sess:\n",
    "    iteration = 1\n",
    "    loss = 0\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "\n",
    "    for e in range(1, epochs+1):\n",
    "        batches = get_batches(train_words, batch_size, window_size)\n",
    "        start = time.time()\n",
    "        # \n",
    "        for x, y in batches:\n",
    "            \n",
    "            feed = {inputs: x,\n",
    "                    labels: np.array(y)[:, None]}\n",
    "            train_loss, _ = sess.run([cost, optimizer], feed_dict=feed)\n",
    "            \n",
    "            loss += train_loss\n",
    "            \n",
    "            if iteration % 100 == 0: \n",
    "                end = time.time()\n",
    "                print(\"Epoch {}/{}\".format(e, epochs),\n",
    "                      \"Iteration: {}\".format(iteration),\n",
    "                      \"Avg. Training loss: {:.4f}\".format(loss/100),\n",
    "                      \"{:.4f} sec/batch\".format((end-start)/100))\n",
    "                loss = 0\n",
    "                start = time.time()\n",
    "            \n",
    "            # 计算相似的词\n",
    "            if iteration % 1000 == 0:\n",
    "                # 计算similarity\n",
    "                sim = similarity.eval()\n",
    "                for i in range(valid_size):\n",
    "                    valid_word = int_to_vocab[valid_examples[i]]\n",
    "                    top_k = 8 # 取最相似单词的前8个\n",
    "                    nearest = (-sim[i, :]).argsort()[1:top_k+1]\n",
    "                    log = 'Nearest to [%s]:' % valid_word\n",
    "                    for k in range(top_k):\n",
    "                        close_word = int_to_vocab[nearest[k]]\n",
    "                        log = '%s %s,' % (log, close_word)\n",
    "                    print(log)\n",
    "            \n",
    "            iteration += 1\n",
    "            \n",
    "#    save_path = saver.save(sess, \"checkpoints/text8.ckpt\")\n",
    "    embed_mat = sess.run(normalized_embedding)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}