data_prepare.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import codecs\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import re\n",
    "#正则表达式"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def origin2tag():\n",
    "    input_data=codecs.open('BosonNLP_NER_6C.txt','r','utf-8')\n",
    "    output_data=codecs.open('./wordtag.txt','w','utf-8')\n",
    "    for line in input_data.readlines():\n",
    "        line=line.strip()\n",
    "        i=0\n",
    "        while i<len(line):\n",
    "            if line[i]=='{':\n",
    "                i+=2\n",
    "                temp=''\n",
    "                while line[i]!='}':\n",
    "                    temp+=line[i]\n",
    "                    i+=1\n",
    "                i+=2\n",
    "                word=temp.split(':')\n",
    "                sen=word[1]\n",
    "                output_data.write(sen[0]+'/B_'+word[0]+' ')\n",
    "                for j in sen[1:len(sen)-1]:\n",
    "                    output_data.write(j+\"/M_\"+word[0]+\" \")\n",
    "                output_data.write(sen[-1]+\"/E_\"+word[0]+\" \")\n",
    "            else:\n",
    "                output_data.write(line[i]+\"/O \")\n",
    "                i+=1\n",
    "        output_data.write('\\n')\n",
    "    input_data.close()\n",
    "    output_data.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#{{product_name:浙江在线杭州}}{{time:4月25日}}讯（记者{{person_name: 施宇翔}} 通讯员 {{person_name:方英}}）毒贩很“时髦”，用{{product_name:微信}}交易毒品。没料想警方也很“潮”，将计就计，一举将其擒获。\n",
    "#浙/B_product_name 江/M_product_name 在/M_product_name 线/M_product_name 杭/M_product_name 州/E_product_name 4/B_time 月/M_time 2/M_time 5/M_time 日/E_time 讯/O （/O 记/O 者/O  /B_person_name 施/M_person_name 宇/M_person_name 翔/E_person_name  /O 通/O 讯/O 员/O  /O 方/B_person_name 英/E_person_name ）/O 毒/O 贩/O 很/O “/O 时/O 髦/O ”/O ，/O 用/O 微/B_product_name 信/E_product_name 交/O 易/O 毒/O 品/O 。/O 没/O 料/O 想/O 警/O 方/O 也/O 很/O “/O 潮/O ”/O ，/O 将/O 计/O 就/O 计/O ，/O 一/O 举/O 将/O 其/O 擒/O 获/O 。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data2pkl():\n",
    "    datas = list()\n",
    "    labels = list()\n",
    "    linedata=list()\n",
    "    linelabel=list()\n",
    "    tags = set()\n",
    "    input_data = codecs.open('./wordtagsplit.txt','r','utf-8')\n",
    "    for line in input_data.readlines():\n",
    "        line = line.split()\n",
    "        linedata=[]\n",
    "        linelabel=[]\n",
    "        num=0\n",
    "        for word in line:\n",
    "            word = word.split('/')\n",
    "            linedata.append(word[0])\n",
    "            linelabel.append(word[1])\n",
    "            tags.add(word[1])\n",
    "            if word[1]!='O':\n",
    "                num=1\n",
    "        if num!=0:\n",
    "            datas.append(linedata)\n",
    "            labels.append(linelabel)\n",
    "    input_data.close()\n",
    "    import collections\n",
    "    def flatten(x):    \n",
    "        result = []    \n",
    "        for el in x:        \n",
    "            if isinstance(x, collections.Iterable) and not isinstance(el, str):            \n",
    "                result.extend(flatten(el))        \n",
    "            else:            \n",
    "                result.append(el)    \n",
    "        return result\n",
    "\n",
    "    \n",
    "    all_words = flatten(datas)\n",
    "    #all_words=list(filter(None,all_words))#去空值\n",
    "    sr_allwords = pd.Series(all_words)\n",
    "    sr_allwords = sr_allwords.value_counts()\n",
    "    set_words = sr_allwords.index\n",
    "    set_ids = range(1, len(set_words)+1)\n",
    "\n",
    "    tags = [i for i in tags]\n",
    "    #\n",
    "    tag_ids = range(len(tags))\n",
    "    word2id = pd.Series(set_ids, index=set_words)\n",
    "    id2word = pd.Series(set_words, index=set_ids)\n",
    "    tag2id = pd.Series(tag_ids, index=tags)\n",
    "    id2tag = pd.Series(tags, index=tag_ids)\n",
    "    word2id[\"unknow\"] = len(word2id)+1\n",
    "    print(len(tags))\n",
    "    print (word2id)\n",
    "    print(tag2id)\n",
    "    #建立一个word2id词典，把每个汉字转换成id。这里习惯现按照数据集中每个汉字出现的次数排序，id从1开始\n",
    "    #再建立一个tag2id词典，把每一个字标注的类型转换成id\n",
    "    \n",
    "    max_len = 60\n",
    "    def X_padding(words):\n",
    "        ids = list(word2id[words])\n",
    "        if len(ids) >= max_len:  \n",
    "            return ids[:max_len]\n",
    "        ids.extend([0]*(max_len-len(ids))) \n",
    "        return ids\n",
    "    def y_padding(tags):\n",
    "        ids = list(tag2id[tags])\n",
    "        if len(ids) >= max_len: \n",
    "            return ids[:max_len]\n",
    "        ids.extend([0]*(max_len-len(ids))) \n",
    "        return ids\n",
    "\n",
    "    df_data = pd.DataFrame({'words': datas, 'tags': labels}, index=range(len(datas)))\n",
    "    df_data['x'] = df_data['words'].apply(X_padding)\n",
    "    df_data['y'] = df_data['tags'].apply(y_padding)\n",
    "    x = np.asarray(list(df_data['x'].values))\n",
    "    y = np.asarray(list(df_data['y'].values))\n",
    "    #print(x)\n",
    "    #print(y)\n",
    "\n",
    "    from sklearn.model_selection import train_test_split\n",
    "    x_train,x_test, y_train, y_test = train_test_split(x, y, test_size=0.2, random_state=43)\n",
    "    x_train, x_valid, y_train, y_valid = train_test_split(x_train, y_train,  test_size=0.2, random_state=43)#训练集抽一部分验证\n",
    "    \n",
    "    \n",
    "    import pickle\n",
    "    import os\n",
    "    with open('Bosondata.pkl', 'wb') as outp:\n",
    "        pickle.dump(word2id, outp)\n",
    "        pickle.dump(id2word, outp)\n",
    "        pickle.dump(tag2id, outp)\n",
    "        pickle.dump(id2tag, outp)\n",
    "        pickle.dump(x_train, outp)\n",
    "        pickle.dump(y_train, outp)\n",
    "        pickle.dump(x_test, outp)\n",
    "        pickle.dump(y_test, outp)\n",
    "        pickle.dump(x_valid, outp)\n",
    "        pickle.dump(y_valid, outp)\n",
    "    print ('** Finished saving the data.')\n",
    "#把刚按标点分开的数据，按照一一对应的顺序，把汉字和每个字的标签转换成id，分别存到两个数组里面，\n",
    "#一起保存到一个pkl文件中，这样模型使用时候就可以直接读取，不用每次都处理数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tagsplit():\n",
    "    with open('wordtag.txt','rb') as inp:\n",
    "        texts=inp.read().decode('utf-8')\n",
    "    sentences=re.split('[，。！？、‘’“”（）]/[O]',texts)\n",
    "    output_data=codecs.open('wordtagsplit.txt','w','utf-8')\n",
    "    for sentence in sentences:\n",
    "        if sentence!='':\n",
    "            output_data.write(sentence.strip()+'\\n')\n",
    "    output_data.close"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "23\n",
      "的            1\n",
      "1            2\n",
      "0            3\n",
      "             4\n",
      "2            5\n",
      "在            6\n",
      "中            7\n",
      "国            8\n",
      "年            9\n",
      "一           10\n",
      "了           11\n",
      "日           12\n",
      "是           13\n",
      "月           14\n",
      "大           15\n",
      "为           16\n",
      "人           17\n",
      "3           18\n",
      "上           19\n",
      "有           20\n",
      "5           21\n",
      "行           22\n",
      "n           23\n",
      "市           24\n",
      "发           25\n",
      "到           26\n",
      "时           27\n",
      "来           28\n",
      "和           29\n",
      "会           30\n",
      "          ... \n",
      "诟         3406\n",
      "辟         3407\n",
      "黯         3408\n",
      "贿         3409\n",
      "屡         3410\n",
      "記         3411\n",
      "捡         3412\n",
      "锂         3413\n",
      "哑         3414\n",
      "锌         3415\n",
      "項         3416\n",
      "窍         3417\n",
      "纺         3418\n",
      "後         3419\n",
      "喇         3420\n",
      "缅         3421\n",
      "堰         3422\n",
      "５         3423\n",
      "斓         3424\n",
      "铠         3425\n",
      "堕         3426\n",
      "鄙         3427\n",
      "琥         3428\n",
      "埂         3429\n",
      "菏         3430\n",
      "崖         3431\n",
      "菁         3432\n",
      "肿         3433\n",
      "侉         3434\n",
      "unknow    3435\n",
      "Length: 3435, dtype: int64\n",
      "E_org_name         0\n",
      "                   1\n",
      "E_ime              2\n",
      "M_company_name     3\n",
      "B_company_name     4\n",
      "E_product_name     5\n",
      "M_org_name         6\n",
      "B_location         7\n",
      "B_time             8\n",
      "M_ime              9\n",
      "M_location        10\n",
      "E_time            11\n",
      "O                 12\n",
      "B_ime             13\n",
      "E_location        14\n",
      "E_company_name    15\n",
      "M_time            16\n",
      "M_person_name     17\n",
      "B_org_name        18\n",
      "M_product_name    19\n",
      "E_person_name     20\n",
      "B_person_name     21\n",
      "B_product_name    22\n",
      "dtype: int64\n",
      "** Finished saving the data.\n"
     ]
    }
   ],
   "source": [
    "#origin2tag()\n",
    "#tagsplit()\n",
    "data2pkl()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>words</th>\n",
       "      <th>tags</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>[浙, 江, 在, 线, 杭, 州, 4, 月, 2, 5, 日, 讯]</td>\n",
       "      <td>[B_product_name, M_product_name, M_product_nam...</td>\n",
       "      <td>[529, 192, 6, 313, 522, 156, 32, 14, 5, 21, 12...</td>\n",
       "      <td>[15, 13, 13, 13, 13, 7, 8, 19, 19, 19, 4, 17, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>[记, 者, , 施, 宇, 翔, , 通, 讯, 员, , 方, 英]</td>\n",
       "      <td>[O, O, B_person_name, M_person_name, M_person_...</td>\n",
       "      <td>[132, 80, 4, 702, 1294, 1582, 4, 111, 184, 134...</td>\n",
       "      <td>[17, 17, 1, 6, 6, 16, 17, 17, 17, 17, 17, 1, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>[用, 微, 信, 交, 易, 毒, 品]</td>\n",
       "      <td>[O, B_product_name, E_product_name, O, O, O, O]</td>\n",
       "      <td>[81, 315, 206, 249, 376, 1274, 101, 0, 0, 0, 0...</td>\n",
       "      <td>[17, 15, 7, 17, 17, 17, 17, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>[记, 者, 从, 杭, 州, 江, 干, 区, 公, 安, 分, 局, 了, 解, 到]</td>\n",
       "      <td>[O, O, O, B_org_name, M_org_name, M_org_name, ...</td>\n",
       "      <td>[132, 80, 160, 522, 156, 192, 874, 109, 38, 14...</td>\n",
       "      <td>[17, 17, 17, 18, 21, 21, 21, 21, 21, 21, 21, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>[江, 干, 区, 禁, 毒, 专, 案, 组, 抓, 获, 吸, 贩, 毒, 人, 员, ...</td>\n",
       "      <td>[B_org_name, M_org_name, M_org_name, M_org_nam...</td>\n",
       "      <td>[192, 874, 109, 1656, 1274, 389, 428, 457, 907...</td>\n",
       "      <td>[18, 21, 21, 21, 21, 21, 21, 11, 17, 17, 17, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>[黑, 龙, 江, 籍, 男, 子, 钱, 某, 长, 期, 落, 脚, 于, 宾, 馆]</td>\n",
       "      <td>[B_location, M_location, E_location, O, O, O, ...</td>\n",
       "      <td>[1055, 558, 192, 1165, 489, 98, 707, 78, 72, 1...</td>\n",
       "      <td>[12, 5, 20, 17, 17, 17, 1, 16, 17, 17, 17, 17,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>[经, 常, 半, 夜, 驾, 车, 来, 往, 于, 杭, 州, 主, 城, 区, 的, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, B_location, M_loca...</td>\n",
       "      <td>[59, 406, 483, 1020, 995, 42, 28, 617, 48, 522...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 12, 5, 5,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>[并, 且, 常, 要, 活, 动, 到, 凌, 晨, 6, 、, 7, 点, 钟]</td>\n",
       "      <td>[O, O, O, O, O, O, O, B_time, M_time, M_time, ...</td>\n",
       "      <td>[251, 841, 406, 120, 356, 66, 26, 1026, 824, 3...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 8, 19, 19, 19, 19...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>[白, 天, 则, 在, 家, 里, 呼, 呼, 大, 睡]</td>\n",
       "      <td>[B_time, E_time, O, O, O, O, O, O, O, O]</td>\n",
       "      <td>[436, 44, 623, 6, 40, 217, 1474, 1474, 15, 206...</td>\n",
       "      <td>[8, 4, 17, 17, 17, 17, 17, 17, 17, 17, 0, 0, 0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>[钱, 某, 不, 寻, 常, 的, 特, 征]</td>\n",
       "      <td>[B_person_name, E_person_name, O, O, O, O, O, O]</td>\n",
       "      <td>[707, 78, 33, 1172, 406, 1, 218, 1197, 0, 0, 0...</td>\n",
       "      <td>[1, 16, 17, 17, 17, 17, 17, 17, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>[发, 现, 钱, 某, 实, 际, 上, 是, 在, 向, 落, 脚, 于, 宾, 馆, ...</td>\n",
       "      <td>[O, O, B_person_name, E_person_name, O, O, O, ...</td>\n",
       "      <td>[25, 85, 707, 78, 235, 448, 19, 13, 6, 255, 83...</td>\n",
       "      <td>[17, 17, 1, 16, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>[在, 即, 将, 举, 办, 的, 第, 十, 一, 届, 广, 州, 国, 际, 汽, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, O, B_location, E_l...</td>\n",
       "      <td>[6, 511, 56, 434, 393, 1, 129, 298, 10, 652, 2...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 12, 2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>[豪, 华, 运, 动, 型, 轿, 车, 全, 新, 英, 菲, 尼, 迪, Q, 5, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, B_product_name, M_...</td>\n",
       "      <td>[1131, 153, 371, 66, 239, 1171, 42, 105, 37, 2...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>[助, 力, 天, 才, 车, 手, 维, 特, 尔, 叱, 咤, 2, 0, 1, 3, ...</td>\n",
       "      <td>[O, O, O, O, O, O, B_person_name, M_person_nam...</td>\n",
       "      <td>[454, 180, 44, 651, 42, 197, 677, 218, 263, 30...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 1, 6, 16, 17, 17, 8, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>[称, 号, 的, 冠, 军, 战, 车, R, B, 9, 也, 将, 震, 撼, 登, 临]</td>\n",
       "      <td>[O, O, O, O, O, O, O, B_product_name, M_produc...</td>\n",
       "      <td>[278, 439, 1, 1086, 225, 296, 42, 452, 299, 31...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 15, 13, 7, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>[创, 造, 2, 0, 1, 3, 收, 视, 率, 奇, 迹, 的, 《, 爸, 爸, ...</td>\n",
       "      <td>[O, O, B_time, M_time, M_time, E_time, O, O, O...</td>\n",
       "      <td>[363, 486, 5, 3, 2, 18, 360, 425, 482, 596, 13...</td>\n",
       "      <td>[17, 17, 8, 19, 19, 4, 17, 17, 17, 17, 17, 17,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>[嘉, 宾, 与, 明, 星, 家, 庭, 专, 属, 座, 驾, 英, 菲, 尼, 迪, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, O, O, B_product_na...</td>\n",
       "      <td>[867, 800, 79, 177, 316, 40, 1007, 389, 742, 9...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>[全, 新, 英, 菲, 尼, 迪, Q, 5, 0, 集, 智, 能]</td>\n",
       "      <td>[O, O, B_product_name, M_product_name, M_produ...</td>\n",
       "      <td>[105, 37, 237, 892, 629, 461, 468, 21, 3, 330,...</td>\n",
       "      <td>[17, 17, 15, 13, 13, 13, 13, 13, 7, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>[搭, 载, 全, 新, 动, 力, 总, 成, 的, 英, 菲, 尼, 迪, Q, 5, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, B_product_name, M_...</td>\n",
       "      <td>[975, 726, 105, 37, 66, 180, 201, 84, 1, 237, ...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>[它, 专, 为, 中, 国, 市, 场, 量, 身, 打, 造]</td>\n",
       "      <td>[O, O, O, B_location, E_location, O, O, O, O, ...</td>\n",
       "      <td>[880, 389, 16, 7, 8, 24, 96, 349, 336, 291, 48...</td>\n",
       "      <td>[17, 17, 17, 12, 20, 17, 17, 17, 17, 17, 17, 0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>[也, 是, 英, 菲, 尼, 迪, 首, 款, 搭, 载, 涡, 轮, 增, 压, 发, ...</td>\n",
       "      <td>[O, O, B_product_name, M_product_name, M_produ...</td>\n",
       "      <td>[71, 13, 237, 892, 629, 461, 264, 173, 975, 72...</td>\n",
       "      <td>[17, 17, 15, 13, 13, 7, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>[英, 菲, 尼, 迪, Q, 5, 0, , 2, ., 0, T, 车, 型, 不, 仅...</td>\n",
       "      <td>[B_product_name, M_product_name, M_product_nam...</td>\n",
       "      <td>[237, 892, 629, 461, 468, 21, 3, 4, 5, 57, 3, ...</td>\n",
       "      <td>[15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 7...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>[将, 助, 推, 英, 菲, 尼, 迪, Q, 5, 0, 成, 为, 同, 级, 别, ...</td>\n",
       "      <td>[O, O, O, B_product_name, M_product_name, M_pr...</td>\n",
       "      <td>[56, 454, 252, 237, 892, 629, 461, 468, 21, 3,...</td>\n",
       "      <td>[17, 17, 17, 15, 13, 13, 13, 13, 13, 7, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>[H, M, C, 证, 券, 分, 析, 师, 格, 雷, 格, ・, 罗]</td>\n",
       "      <td>[B_company_name, M_company_name, E_company_nam...</td>\n",
       "      <td>[552, 215, 152, 386, 938, 75, 890, 395, 401, 7...</td>\n",
       "      <td>[2, 9, 14, 17, 17, 17, 17, 17, 17, 1, 6, 6, 16...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>[G, r, e, g, , R, o, h]</td>\n",
       "      <td>[B_person_name, M_person_name, M_person_name, ...</td>\n",
       "      <td>[246, 100, 52, 516, 4, 452, 47, 219, 0, 0, 0, ...</td>\n",
       "      <td>[1, 6, 6, 6, 6, 6, 6, 16, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>[L, G, 手, 机, 的, 质, 量, 有, 了, 一, 定, 的, 提, 升]</td>\n",
       "      <td>[B_company_name, E_company_name, O, O, O, O, O...</td>\n",
       "      <td>[265, 246, 197, 83, 1, 637, 349, 20, 11, 10, 2...</td>\n",
       "      <td>[2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>[L, G, 与, 三, 星]</td>\n",
       "      <td>[B_company_name, E_company_name, O, B_company_...</td>\n",
       "      <td>[265, 246, 79, 122, 316, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>[2, 14, 17, 2, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>[苹, 果, 等, 手, 机, 巨, 头, 仍, 有, 不, 小, 的, 差, 距]</td>\n",
       "      <td>[B_company_name, E_company_name, O, O, O, O, O...</td>\n",
       "      <td>[843, 372, 159, 197, 83, 769, 500, 784, 20, 33...</td>\n",
       "      <td>[2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>[L, G, 手, 机, 在, 工, 业, 设, 计, 及, 产, 品, 本, 身, 以, ...</td>\n",
       "      <td>[B_company_name, E_company_name, O, O, O, O, O...</td>\n",
       "      <td>[265, 246, 197, 83, 6, 93, 43, 262, 240, 221, ...</td>\n",
       "      <td>[2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>[产, 品, 的, 细, 节, 化, 始, 终, 没, 有, 三, 星, 完, 善]</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, O, B_company_name,...</td>\n",
       "      <td>[113, 101, 1, 1189, 458, 242, 375, 668, 383, 2...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 2, 14...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16723</th>\n",
       "      <td>[无, 锡, 绿, 城, 总, 经, 理, 顾, 国, 华, 先, 生, 上, 台, 致, ...</td>\n",
       "      <td>[B_company_name, M_company_name, M_company_nam...</td>\n",
       "      <td>[334, 1128, 1319, 118, 201, 59, 133, 773, 8, 1...</td>\n",
       "      <td>[2, 9, 9, 14, 17, 17, 17, 1, 6, 16, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16724</th>\n",
       "      <td>[无, 锡, 绿, 城, 置, 业, 副, 总, 经, 理, 华, 超, 先, 生, 发, ...</td>\n",
       "      <td>[B_company_name, M_company_name, M_company_nam...</td>\n",
       "      <td>[334, 1128, 1319, 118, 687, 43, 492, 201, 59, ...</td>\n",
       "      <td>[2, 9, 9, 14, 17, 17, 17, 17, 17, 17, 1, 16, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16725</th>\n",
       "      <td>[向, 到, 场, 嘉, 宾, 介, 绍, 绿, 城, 玉, 兰, 花, 园, 高, 端, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, B_product_name, M_produc...</td>\n",
       "      <td>[255, 26, 96, 867, 800, 227, 615, 1319, 118, 7...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 15, 13, 13, 13, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16726</th>\n",
       "      <td>[玉, 兰, 会, 为, 业, 主, 提, 供, 全, 方, 位, 服, 务]</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[720, 589, 30, 16, 43, 94, 193, 438, 105, 99, ...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16727</th>\n",
       "      <td>[车, P, a, d, Ⅱ, 各, 项, 配, 置, 和, 新, 增, 功, 能, 值, ...</td>\n",
       "      <td>[B_product_name, M_product_name, M_product_nam...</td>\n",
       "      <td>[42, 171, 61, 245, 2251, 480, 413, 499, 687, 2...</td>\n",
       "      <td>[15, 13, 13, 13, 7, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16728</th>\n",
       "      <td>[目, 前, 车, P, a, d, Ⅱ, 的, 价, 格, 还, 是, 一, 个, 谜]</td>\n",
       "      <td>[O, O, B_product_name, M_product_name, M_produ...</td>\n",
       "      <td>[234, 62, 42, 171, 61, 245, 2251, 1, 328, 401,...</td>\n",
       "      <td>[17, 17, 15, 13, 13, 13, 7, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16729</th>\n",
       "      <td>[银, 泰, 网, 力, 求, 迎, 合, 品, 牌, 商, 的, 品, 位]</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[158, 894, 34, 180, 565, 728, 139, 101, 405, 7...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16730</th>\n",
       "      <td>[使, 他, 们, 感, 觉, 到, 银, 泰, 网, 的, 品, 牌, 形, 象, 与, ...</td>\n",
       "      <td>[O, O, O, O, O, O, B_product_name, M_product_n...</td>\n",
       "      <td>[373, 157, 168, 539, 1102, 26, 158, 894, 34, 1...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 15, 13, 7, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16731</th>\n",
       "      <td>[而, 银, 泰, 网, 对, 于, 服, 务, 自, 然, 有, 完, 整, 的, 体, 系]</td>\n",
       "      <td>[O, B_product_name, M_product_name, E_product_...</td>\n",
       "      <td>[138, 158, 894, 34, 55, 48, 415, 110, 107, 355...</td>\n",
       "      <td>[17, 15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16732</th>\n",
       "      <td>[若, 产, 品, 有, 问, 题, 投, 诉, 银, 泰, 网, 则, 会, 第, 一, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, B_product_name, M_pro...</td>\n",
       "      <td>[1782, 113, 101, 20, 348, 446, 297, 598, 158, ...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 7, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16733</th>\n",
       "      <td>[用, 廖, 斌, 的, 话, 说]</td>\n",
       "      <td>[O, B_person_name, E_person_name, O, O, O]</td>\n",
       "      <td>[81, 1941, 1720, 1, 481, 114, 0, 0, 0, 0, 0, 0...</td>\n",
       "      <td>[17, 1, 16, 17, 17, 17, 0, 0, 0, 0, 0, 0, 0, 0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16734</th>\n",
       "      <td>[在, 廖, 斌, 看, 来]</td>\n",
       "      <td>[O, B_person_name, E_person_name, O, O]</td>\n",
       "      <td>[6, 1941, 1720, 267, 28, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>[17, 1, 16, 17, 17, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16735</th>\n",
       "      <td>[每, 家, 银, 泰, 百, 货, 都, 会, 是, 银, 泰, 网, 最, 好, 的, ...</td>\n",
       "      <td>[O, O, B_company_name, M_company_name, M_compa...</td>\n",
       "      <td>[568, 40, 158, 894, 466, 654, 161, 30, 13, 158...</td>\n",
       "      <td>[17, 17, 2, 9, 9, 14, 17, 17, 17, 15, 13, 7, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16736</th>\n",
       "      <td>[而, 银, 泰, 百, 货, 也, 逐, 渐, 成, 为, 银, 泰, 网, 的]</td>\n",
       "      <td>[O, B_company_name, M_company_name, M_company_...</td>\n",
       "      <td>[138, 158, 894, 466, 654, 71, 1174, 1618, 84, ...</td>\n",
       "      <td>[17, 2, 9, 9, 14, 17, 17, 17, 17, 17, 15, 13, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16737</th>\n",
       "      <td>[8, 月, 1, 5, 日]</td>\n",
       "      <td>[B_time, M_time, M_time, M_time, E_time]</td>\n",
       "      <td>[41, 14, 2, 21, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "      <td>[8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16738</th>\n",
       "      <td>[高, 朋, 网, 被, 内, 部, 员, 工, 爆, 料]</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[91, 1068, 34, 182, 124, 95, 134, 93, 1215, 90...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16739</th>\n",
       "      <td>[在, 全, 国, 1, 2, 8, 个, 城, 市, 开, 了, 分, 站, 的, 知, ...</td>\n",
       "      <td>[O, O, O, O, O, O, O, O, O, O, O, O, O, O, O, ...</td>\n",
       "      <td>[6, 105, 8, 2, 5, 41, 35, 118, 24, 63, 11, 75,...</td>\n",
       "      <td>[17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16740</th>\n",
       "      <td>[8, 月, 1, 8, 日]</td>\n",
       "      <td>[B_time, M_time, M_time, M_time, E_time]</td>\n",
       "      <td>[41, 14, 2, 41, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "      <td>[8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16741</th>\n",
       "      <td>[7, 月, 份, 团, 购, 业, 的, 月, 销, 售, 总, 额, 首, 次, 突, ...</td>\n",
       "      <td>[B_time, E_time, O, O, O, O, O, O, O, O, O, O,...</td>\n",
       "      <td>[46, 14, 370, 238, 211, 43, 1, 14, 294, 341, 2...</td>\n",
       "      <td>[8, 4, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16742</th>\n",
       "      <td>[其, 中, 拉, 手, 网]</td>\n",
       "      <td>[O, O, B_product_name, M_product_name, E_produ...</td>\n",
       "      <td>[189, 7, 420, 197, 34, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "      <td>[17, 17, 15, 13, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16743</th>\n",
       "      <td>[窝, 窝, 团]</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_name]</td>\n",
       "      <td>[1183, 1183, 238, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0...</td>\n",
       "      <td>[15, 13, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16744</th>\n",
       "      <td>[美, 团, 网, 等, 5, 家, 团, 购, 网, 站, 销, 售, 额, 分, 别, ...</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[45, 238, 34, 159, 21, 40, 238, 211, 34, 447, ...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16745</th>\n",
       "      <td>[8, 月, 1, 9, 日]</td>\n",
       "      <td>[B_time, M_time, M_time, M_time, E_time]</td>\n",
       "      <td>[41, 14, 2, 31, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...</td>\n",
       "      <td>[8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16746</th>\n",
       "      <td>[2, 0, 1, 1, 年, 8, 月, 1, 9, 日, 由, 中, 国, 电, 子, ...</td>\n",
       "      <td>[B_time, M_time, M_time, M_time, M_time, M_tim...</td>\n",
       "      <td>[5, 3, 2, 2, 9, 41, 14, 2, 31, 12, 279, 7, 8, ...</td>\n",
       "      <td>[8, 19, 19, 19, 19, 19, 19, 19, 19, 4, 17, 18,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16747</th>\n",
       "      <td>[中, 国, 电, 子, 商, 务, 投, 诉, 与, 维, 权, 公, 共, 服, 务, ...</td>\n",
       "      <td>[B_product_name, M_product_name, M_product_nam...</td>\n",
       "      <td>[7, 8, 77, 98, 76, 110, 297, 598, 79, 677, 433...</td>\n",
       "      <td>[15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16748</th>\n",
       "      <td>[(, w, w, w, ., 1, 0, 0, e, c, ., c, n, , z, t...</td>\n",
       "      <td>[B_product_name, M_product_name, M_product_nam...</td>\n",
       "      <td>[130, 502, 502, 502, 57, 2, 3, 3, 52, 209, 57,...</td>\n",
       "      <td>[15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16749</th>\n",
       "      <td>[赛, 飞, 利, 作, 为, 欧, 宝, 在, 欧, 洲, 最, 畅, 销, 的, 多, ...</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[377, 656, 226, 73, 16, 494, 144, 6, 494, 634,...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 2, 14, 17, 12, 20, 17, 17,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16750</th>\n",
       "      <td>[使, 赛, 飞, 利, 的, 整, 体, 品, 味, 都, 有, 较, 大, 的, 提, 升]</td>\n",
       "      <td>[O, B_product_name, M_product_name, E_product_...</td>\n",
       "      <td>[373, 377, 656, 226, 1, 540, 212, 101, 1326, 1...</td>\n",
       "      <td>[17, 15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16751</th>\n",
       "      <td>[动, 力, 上, 赛, 飞, 利, 搭, 载, 了, E, C, O, T, E, C, ...</td>\n",
       "      <td>[O, O, O, B_product_name, M_product_name, E_pr...</td>\n",
       "      <td>[66, 180, 19, 377, 656, 226, 975, 726, 11, 300...</td>\n",
       "      <td>[17, 17, 17, 15, 13, 7, 17, 17, 17, 17, 17, 17...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16752</th>\n",
       "      <td>[赛, 飞, 利, 配, 置, 了, A, B, S, , B, A, , C, B, C,...</td>\n",
       "      <td>[B_product_name, M_product_name, E_product_nam...</td>\n",
       "      <td>[377, 656, 226, 499, 687, 11, 127, 299, 147, 4...</td>\n",
       "      <td>[15, 13, 7, 17, 17, 17, 17, 17, 17, 0, 17, 17,...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>16753 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   words  \\\n",
       "0                   [浙, 江, 在, 线, 杭, 州, 4, 月, 2, 5, 日, 讯]   \n",
       "1                   [记, 者, , 施, 宇, 翔, , 通, 讯, 员, , 方, 英]   \n",
       "2                                  [用, 微, 信, 交, 易, 毒, 品]   \n",
       "3          [记, 者, 从, 杭, 州, 江, 干, 区, 公, 安, 分, 局, 了, 解, 到]   \n",
       "4      [江, 干, 区, 禁, 毒, 专, 案, 组, 抓, 获, 吸, 贩, 毒, 人, 员, ...   \n",
       "5          [黑, 龙, 江, 籍, 男, 子, 钱, 某, 长, 期, 落, 脚, 于, 宾, 馆]   \n",
       "6      [经, 常, 半, 夜, 驾, 车, 来, 往, 于, 杭, 州, 主, 城, 区, 的, ...   \n",
       "7             [并, 且, 常, 要, 活, 动, 到, 凌, 晨, 6, 、, 7, 点, 钟]   \n",
       "8                         [白, 天, 则, 在, 家, 里, 呼, 呼, 大, 睡]   \n",
       "9                               [钱, 某, 不, 寻, 常, 的, 特, 征]   \n",
       "10     [发, 现, 钱, 某, 实, 际, 上, 是, 在, 向, 落, 脚, 于, 宾, 馆, ...   \n",
       "11     [在, 即, 将, 举, 办, 的, 第, 十, 一, 届, 广, 州, 国, 际, 汽, ...   \n",
       "12     [豪, 华, 运, 动, 型, 轿, 车, 全, 新, 英, 菲, 尼, 迪, Q, 5, ...   \n",
       "13     [助, 力, 天, 才, 车, 手, 维, 特, 尔, 叱, 咤, 2, 0, 1, 3, ...   \n",
       "14      [称, 号, 的, 冠, 军, 战, 车, R, B, 9, 也, 将, 震, 撼, 登, 临]   \n",
       "15     [创, 造, 2, 0, 1, 3, 收, 视, 率, 奇, 迹, 的, 《, 爸, 爸, ...   \n",
       "16     [嘉, 宾, 与, 明, 星, 家, 庭, 专, 属, 座, 驾, 英, 菲, 尼, 迪, ...   \n",
       "17                  [全, 新, 英, 菲, 尼, 迪, Q, 5, 0, 集, 智, 能]   \n",
       "18     [搭, 载, 全, 新, 动, 力, 总, 成, 的, 英, 菲, 尼, 迪, Q, 5, ...   \n",
       "19                     [它, 专, 为, 中, 国, 市, 场, 量, 身, 打, 造]   \n",
       "20     [也, 是, 英, 菲, 尼, 迪, 首, 款, 搭, 载, 涡, 轮, 增, 压, 发, ...   \n",
       "21     [英, 菲, 尼, 迪, Q, 5, 0, , 2, ., 0, T, 车, 型, 不, 仅...   \n",
       "22     [将, 助, 推, 英, 菲, 尼, 迪, Q, 5, 0, 成, 为, 同, 级, 别, ...   \n",
       "23               [H, M, C, 证, 券, 分, 析, 师, 格, 雷, 格, ・, 罗]   \n",
       "24                               [G, r, e, g, , R, o, h]   \n",
       "25            [L, G, 手, 机, 的, 质, 量, 有, 了, 一, 定, 的, 提, 升]   \n",
       "26                                       [L, G, 与, 三, 星]   \n",
       "27            [苹, 果, 等, 手, 机, 巨, 头, 仍, 有, 不, 小, 的, 差, 距]   \n",
       "28     [L, G, 手, 机, 在, 工, 业, 设, 计, 及, 产, 品, 本, 身, 以, ...   \n",
       "29            [产, 品, 的, 细, 节, 化, 始, 终, 没, 有, 三, 星, 完, 善]   \n",
       "...                                                  ...   \n",
       "16723  [无, 锡, 绿, 城, 总, 经, 理, 顾, 国, 华, 先, 生, 上, 台, 致, ...   \n",
       "16724  [无, 锡, 绿, 城, 置, 业, 副, 总, 经, 理, 华, 超, 先, 生, 发, ...   \n",
       "16725  [向, 到, 场, 嘉, 宾, 介, 绍, 绿, 城, 玉, 兰, 花, 园, 高, 端, ...   \n",
       "16726            [玉, 兰, 会, 为, 业, 主, 提, 供, 全, 方, 位, 服, 务]   \n",
       "16727  [车, P, a, d, Ⅱ, 各, 项, 配, 置, 和, 新, 增, 功, 能, 值, ...   \n",
       "16728      [目, 前, 车, P, a, d, Ⅱ, 的, 价, 格, 还, 是, 一, 个, 谜]   \n",
       "16729            [银, 泰, 网, 力, 求, 迎, 合, 品, 牌, 商, 的, 品, 位]   \n",
       "16730  [使, 他, 们, 感, 觉, 到, 银, 泰, 网, 的, 品, 牌, 形, 象, 与, ...   \n",
       "16731   [而, 银, 泰, 网, 对, 于, 服, 务, 自, 然, 有, 完, 整, 的, 体, 系]   \n",
       "16732  [若, 产, 品, 有, 问, 题, 投, 诉, 银, 泰, 网, 则, 会, 第, 一, ...   \n",
       "16733                                 [用, 廖, 斌, 的, 话, 说]   \n",
       "16734                                    [在, 廖, 斌, 看, 来]   \n",
       "16735  [每, 家, 银, 泰, 百, 货, 都, 会, 是, 银, 泰, 网, 最, 好, 的, ...   \n",
       "16736         [而, 银, 泰, 百, 货, 也, 逐, 渐, 成, 为, 银, 泰, 网, 的]   \n",
       "16737                                    [8, 月, 1, 5, 日]   \n",
       "16738                     [高, 朋, 网, 被, 内, 部, 员, 工, 爆, 料]   \n",
       "16739  [在, 全, 国, 1, 2, 8, 个, 城, 市, 开, 了, 分, 站, 的, 知, ...   \n",
       "16740                                    [8, 月, 1, 8, 日]   \n",
       "16741  [7, 月, 份, 团, 购, 业, 的, 月, 销, 售, 总, 额, 首, 次, 突, ...   \n",
       "16742                                    [其, 中, 拉, 手, 网]   \n",
       "16743                                          [窝, 窝, 团]   \n",
       "16744  [美, 团, 网, 等, 5, 家, 团, 购, 网, 站, 销, 售, 额, 分, 别, ...   \n",
       "16745                                    [8, 月, 1, 9, 日]   \n",
       "16746  [2, 0, 1, 1, 年, 8, 月, 1, 9, 日, 由, 中, 国, 电, 子, ...   \n",
       "16747  [中, 国, 电, 子, 商, 务, 投, 诉, 与, 维, 权, 公, 共, 服, 务, ...   \n",
       "16748  [(, w, w, w, ., 1, 0, 0, e, c, ., c, n, , z, t...   \n",
       "16749  [赛, 飞, 利, 作, 为, 欧, 宝, 在, 欧, 洲, 最, 畅, 销, 的, 多, ...   \n",
       "16750   [使, 赛, 飞, 利, 的, 整, 体, 品, 味, 都, 有, 较, 大, 的, 提, 升]   \n",
       "16751  [动, 力, 上, 赛, 飞, 利, 搭, 载, 了, E, C, O, T, E, C, ...   \n",
       "16752  [赛, 飞, 利, 配, 置, 了, A, B, S, , B, A, , C, B, C,...   \n",
       "\n",
       "                                                    tags  \\\n",
       "0      [B_product_name, M_product_name, M_product_nam...   \n",
       "1      [O, O, B_person_name, M_person_name, M_person_...   \n",
       "2        [O, B_product_name, E_product_name, O, O, O, O]   \n",
       "3      [O, O, O, B_org_name, M_org_name, M_org_name, ...   \n",
       "4      [B_org_name, M_org_name, M_org_name, M_org_nam...   \n",
       "5      [B_location, M_location, E_location, O, O, O, ...   \n",
       "6      [O, O, O, O, O, O, O, O, O, B_location, M_loca...   \n",
       "7      [O, O, O, O, O, O, O, B_time, M_time, M_time, ...   \n",
       "8               [B_time, E_time, O, O, O, O, O, O, O, O]   \n",
       "9       [B_person_name, E_person_name, O, O, O, O, O, O]   \n",
       "10     [O, O, B_person_name, E_person_name, O, O, O, ...   \n",
       "11     [O, O, O, O, O, O, O, O, O, O, B_location, E_l...   \n",
       "12     [O, O, O, O, O, O, O, O, O, B_product_name, M_...   \n",
       "13     [O, O, O, O, O, O, B_person_name, M_person_nam...   \n",
       "14     [O, O, O, O, O, O, O, B_product_name, M_produc...   \n",
       "15     [O, O, B_time, M_time, M_time, E_time, O, O, O...   \n",
       "16     [O, O, O, O, O, O, O, O, O, O, O, B_product_na...   \n",
       "17     [O, O, B_product_name, M_product_name, M_produ...   \n",
       "18     [O, O, O, O, O, O, O, O, O, B_product_name, M_...   \n",
       "19     [O, O, O, B_location, E_location, O, O, O, O, ...   \n",
       "20     [O, O, B_product_name, M_product_name, M_produ...   \n",
       "21     [B_product_name, M_product_name, M_product_nam...   \n",
       "22     [O, O, O, B_product_name, M_product_name, M_pr...   \n",
       "23     [B_company_name, M_company_name, E_company_nam...   \n",
       "24     [B_person_name, M_person_name, M_person_name, ...   \n",
       "25     [B_company_name, E_company_name, O, O, O, O, O...   \n",
       "26     [B_company_name, E_company_name, O, B_company_...   \n",
       "27     [B_company_name, E_company_name, O, O, O, O, O...   \n",
       "28     [B_company_name, E_company_name, O, O, O, O, O...   \n",
       "29     [O, O, O, O, O, O, O, O, O, O, B_company_name,...   \n",
       "...                                                  ...   \n",
       "16723  [B_company_name, M_company_name, M_company_nam...   \n",
       "16724  [B_company_name, M_company_name, M_company_nam...   \n",
       "16725  [O, O, O, O, O, O, O, B_product_name, M_produc...   \n",
       "16726  [B_product_name, M_product_name, E_product_nam...   \n",
       "16727  [B_product_name, M_product_name, M_product_nam...   \n",
       "16728  [O, O, B_product_name, M_product_name, M_produ...   \n",
       "16729  [B_product_name, M_product_name, E_product_nam...   \n",
       "16730  [O, O, O, O, O, O, B_product_name, M_product_n...   \n",
       "16731  [O, B_product_name, M_product_name, E_product_...   \n",
       "16732  [O, O, O, O, O, O, O, O, B_product_name, M_pro...   \n",
       "16733         [O, B_person_name, E_person_name, O, O, O]   \n",
       "16734            [O, B_person_name, E_person_name, O, O]   \n",
       "16735  [O, O, B_company_name, M_company_name, M_compa...   \n",
       "16736  [O, B_company_name, M_company_name, M_company_...   \n",
       "16737           [B_time, M_time, M_time, M_time, E_time]   \n",
       "16738  [B_product_name, M_product_name, E_product_nam...   \n",
       "16739  [O, O, O, O, O, O, O, O, O, O, O, O, O, O, O, ...   \n",
       "16740           [B_time, M_time, M_time, M_time, E_time]   \n",
       "16741  [B_time, E_time, O, O, O, O, O, O, O, O, O, O,...   \n",
       "16742  [O, O, B_product_name, M_product_name, E_produ...   \n",
       "16743   [B_product_name, M_product_name, E_product_name]   \n",
       "16744  [B_product_name, M_product_name, E_product_nam...   \n",
       "16745           [B_time, M_time, M_time, M_time, E_time]   \n",
       "16746  [B_time, M_time, M_time, M_time, M_time, M_tim...   \n",
       "16747  [B_product_name, M_product_name, M_product_nam...   \n",
       "16748  [B_product_name, M_product_name, M_product_nam...   \n",
       "16749  [B_product_name, M_product_name, E_product_nam...   \n",
       "16750  [O, B_product_name, M_product_name, E_product_...   \n",
       "16751  [O, O, O, B_product_name, M_product_name, E_pr...   \n",
       "16752  [B_product_name, M_product_name, E_product_nam...   \n",
       "\n",
       "                                                       x  \\\n",
       "0      [529, 192, 6, 313, 522, 156, 32, 14, 5, 21, 12...   \n",
       "1      [132, 80, 4, 702, 1294, 1582, 4, 111, 184, 134...   \n",
       "2      [81, 315, 206, 249, 376, 1274, 101, 0, 0, 0, 0...   \n",
       "3      [132, 80, 160, 522, 156, 192, 874, 109, 38, 14...   \n",
       "4      [192, 874, 109, 1656, 1274, 389, 428, 457, 907...   \n",
       "5      [1055, 558, 192, 1165, 489, 98, 707, 78, 72, 1...   \n",
       "6      [59, 406, 483, 1020, 995, 42, 28, 617, 48, 522...   \n",
       "7      [251, 841, 406, 120, 356, 66, 26, 1026, 824, 3...   \n",
       "8      [436, 44, 623, 6, 40, 217, 1474, 1474, 15, 206...   \n",
       "9      [707, 78, 33, 1172, 406, 1, 218, 1197, 0, 0, 0...   \n",
       "10     [25, 85, 707, 78, 235, 448, 19, 13, 6, 255, 83...   \n",
       "11     [6, 511, 56, 434, 393, 1, 129, 298, 10, 652, 2...   \n",
       "12     [1131, 153, 371, 66, 239, 1171, 42, 105, 37, 2...   \n",
       "13     [454, 180, 44, 651, 42, 197, 677, 218, 263, 30...   \n",
       "14     [278, 439, 1, 1086, 225, 296, 42, 452, 299, 31...   \n",
       "15     [363, 486, 5, 3, 2, 18, 360, 425, 482, 596, 13...   \n",
       "16     [867, 800, 79, 177, 316, 40, 1007, 389, 742, 9...   \n",
       "17     [105, 37, 237, 892, 629, 461, 468, 21, 3, 330,...   \n",
       "18     [975, 726, 105, 37, 66, 180, 201, 84, 1, 237, ...   \n",
       "19     [880, 389, 16, 7, 8, 24, 96, 349, 336, 291, 48...   \n",
       "20     [71, 13, 237, 892, 629, 461, 264, 173, 975, 72...   \n",
       "21     [237, 892, 629, 461, 468, 21, 3, 4, 5, 57, 3, ...   \n",
       "22     [56, 454, 252, 237, 892, 629, 461, 468, 21, 3,...   \n",
       "23     [552, 215, 152, 386, 938, 75, 890, 395, 401, 7...   \n",
       "24     [246, 100, 52, 516, 4, 452, 47, 219, 0, 0, 0, ...   \n",
       "25     [265, 246, 197, 83, 1, 637, 349, 20, 11, 10, 2...   \n",
       "26     [265, 246, 79, 122, 316, 0, 0, 0, 0, 0, 0, 0, ...   \n",
       "27     [843, 372, 159, 197, 83, 769, 500, 784, 20, 33...   \n",
       "28     [265, 246, 197, 83, 6, 93, 43, 262, 240, 221, ...   \n",
       "29     [113, 101, 1, 1189, 458, 242, 375, 668, 383, 2...   \n",
       "...                                                  ...   \n",
       "16723  [334, 1128, 1319, 118, 201, 59, 133, 773, 8, 1...   \n",
       "16724  [334, 1128, 1319, 118, 687, 43, 492, 201, 59, ...   \n",
       "16725  [255, 26, 96, 867, 800, 227, 615, 1319, 118, 7...   \n",
       "16726  [720, 589, 30, 16, 43, 94, 193, 438, 105, 99, ...   \n",
       "16727  [42, 171, 61, 245, 2251, 480, 413, 499, 687, 2...   \n",
       "16728  [234, 62, 42, 171, 61, 245, 2251, 1, 328, 401,...   \n",
       "16729  [158, 894, 34, 180, 565, 728, 139, 101, 405, 7...   \n",
       "16730  [373, 157, 168, 539, 1102, 26, 158, 894, 34, 1...   \n",
       "16731  [138, 158, 894, 34, 55, 48, 415, 110, 107, 355...   \n",
       "16732  [1782, 113, 101, 20, 348, 446, 297, 598, 158, ...   \n",
       "16733  [81, 1941, 1720, 1, 481, 114, 0, 0, 0, 0, 0, 0...   \n",
       "16734  [6, 1941, 1720, 267, 28, 0, 0, 0, 0, 0, 0, 0, ...   \n",
       "16735  [568, 40, 158, 894, 466, 654, 161, 30, 13, 158...   \n",
       "16736  [138, 158, 894, 466, 654, 71, 1174, 1618, 84, ...   \n",
       "16737  [41, 14, 2, 21, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
       "16738  [91, 1068, 34, 182, 124, 95, 134, 93, 1215, 90...   \n",
       "16739  [6, 105, 8, 2, 5, 41, 35, 118, 24, 63, 11, 75,...   \n",
       "16740  [41, 14, 2, 41, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
       "16741  [46, 14, 370, 238, 211, 43, 1, 14, 294, 341, 2...   \n",
       "16742  [189, 7, 420, 197, 34, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
       "16743  [1183, 1183, 238, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0...   \n",
       "16744  [45, 238, 34, 159, 21, 40, 238, 211, 34, 447, ...   \n",
       "16745  [41, 14, 2, 31, 12, 0, 0, 0, 0, 0, 0, 0, 0, 0,...   \n",
       "16746  [5, 3, 2, 2, 9, 41, 14, 2, 31, 12, 279, 7, 8, ...   \n",
       "16747  [7, 8, 77, 98, 76, 110, 297, 598, 79, 677, 433...   \n",
       "16748  [130, 502, 502, 502, 57, 2, 3, 3, 52, 209, 57,...   \n",
       "16749  [377, 656, 226, 73, 16, 494, 144, 6, 494, 634,...   \n",
       "16750  [373, 377, 656, 226, 1, 540, 212, 101, 1326, 1...   \n",
       "16751  [66, 180, 19, 377, 656, 226, 975, 726, 11, 300...   \n",
       "16752  [377, 656, 226, 499, 687, 11, 127, 299, 147, 4...   \n",
       "\n",
       "                                                       y  \n",
       "0      [15, 13, 13, 13, 13, 7, 8, 19, 19, 19, 4, 17, ...  \n",
       "1      [17, 17, 1, 6, 6, 16, 17, 17, 17, 17, 17, 1, 1...  \n",
       "2      [17, 15, 7, 17, 17, 17, 17, 0, 0, 0, 0, 0, 0, ...  \n",
       "3      [17, 17, 17, 18, 21, 21, 21, 21, 21, 21, 21, 1...  \n",
       "4      [18, 21, 21, 21, 21, 21, 21, 11, 17, 17, 17, 1...  \n",
       "5      [12, 5, 20, 17, 17, 17, 1, 16, 17, 17, 17, 17,...  \n",
       "6      [17, 17, 17, 17, 17, 17, 17, 17, 17, 12, 5, 5,...  \n",
       "7      [17, 17, 17, 17, 17, 17, 17, 8, 19, 19, 19, 19...  \n",
       "8      [8, 4, 17, 17, 17, 17, 17, 17, 17, 17, 0, 0, 0...  \n",
       "9      [1, 16, 17, 17, 17, 17, 17, 17, 0, 0, 0, 0, 0,...  \n",
       "10     [17, 17, 1, 16, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "11     [17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 12, 2...  \n",
       "12     [17, 17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 1...  \n",
       "13     [17, 17, 17, 17, 17, 17, 1, 6, 16, 17, 17, 8, ...  \n",
       "14     [17, 17, 17, 17, 17, 17, 17, 15, 13, 7, 17, 17...  \n",
       "15     [17, 17, 8, 19, 19, 4, 17, 17, 17, 17, 17, 17,...  \n",
       "16     [17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 1...  \n",
       "17     [17, 17, 15, 13, 13, 13, 13, 13, 7, 17, 17, 17...  \n",
       "18     [17, 17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 1...  \n",
       "19     [17, 17, 17, 12, 20, 17, 17, 17, 17, 17, 17, 0...  \n",
       "20     [17, 17, 15, 13, 13, 7, 17, 17, 17, 17, 17, 17...  \n",
       "21     [15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 7...  \n",
       "22     [17, 17, 17, 15, 13, 13, 13, 13, 13, 7, 17, 17...  \n",
       "23     [2, 9, 14, 17, 17, 17, 17, 17, 17, 1, 6, 6, 16...  \n",
       "24     [1, 6, 6, 6, 6, 6, 6, 16, 0, 0, 0, 0, 0, 0, 0,...  \n",
       "25     [2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "26     [2, 14, 17, 2, 14, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
       "27     [2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "28     [2, 14, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "29     [17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 2, 14...  \n",
       "...                                                  ...  \n",
       "16723  [2, 9, 9, 14, 17, 17, 17, 1, 6, 16, 17, 17, 17...  \n",
       "16724  [2, 9, 9, 14, 17, 17, 17, 17, 17, 17, 1, 16, 1...  \n",
       "16725  [17, 17, 17, 17, 17, 17, 17, 15, 13, 13, 13, 1...  \n",
       "16726  [15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16727  [15, 13, 13, 13, 7, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16728  [17, 17, 15, 13, 13, 13, 7, 17, 17, 17, 17, 17...  \n",
       "16729  [15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16730  [17, 17, 17, 17, 17, 17, 15, 13, 7, 17, 17, 17...  \n",
       "16731  [17, 15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16732  [17, 17, 17, 17, 17, 17, 17, 17, 15, 13, 7, 17...  \n",
       "16733  [17, 1, 16, 17, 17, 17, 0, 0, 0, 0, 0, 0, 0, 0...  \n",
       "16734  [17, 1, 16, 17, 17, 0, 0, 0, 0, 0, 0, 0, 0, 0,...  \n",
       "16735  [17, 17, 2, 9, 9, 14, 17, 17, 17, 15, 13, 7, 1...  \n",
       "16736  [17, 2, 9, 9, 14, 17, 17, 17, 17, 17, 15, 13, ...  \n",
       "16737  [8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
       "16738  [15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 0, 0, ...  \n",
       "16739  [17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 1...  \n",
       "16740  [8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
       "16741  [8, 4, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17,...  \n",
       "16742  [17, 17, 15, 13, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0,...  \n",
       "16743  [15, 13, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0...  \n",
       "16744  [15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16745  [8, 19, 19, 19, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...  \n",
       "16746  [8, 19, 19, 19, 19, 19, 19, 19, 19, 4, 17, 18,...  \n",
       "16747  [15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 1...  \n",
       "16748  [15, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 1...  \n",
       "16749  [15, 13, 7, 17, 17, 2, 14, 17, 12, 20, 17, 17,...  \n",
       "16750  [17, 15, 13, 7, 17, 17, 17, 17, 17, 17, 17, 17...  \n",
       "16751  [17, 17, 17, 15, 13, 7, 17, 17, 17, 17, 17, 17...  \n",
       "16752  [15, 13, 7, 17, 17, 17, 17, 17, 17, 0, 17, 17,...  \n",
       "\n",
       "[16753 rows x 4 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}