2017-7-23 论文数据库资料.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "data_preparation/retreive_news.py:232: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  df_title[\"head_tagged_jp\"] = df_title[\"HEADLINE_ALERT_TEXT_x\"].apply(tagging_jp)\n",
      "data_preparation/retreive_news.py:233: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  df_title[\"head_tagged2_jp\"] = df_title[\"head_tagged_jp\"].apply(clean_tag_jp)\n",
      "data_preparation/retreive_news.py:236: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  df_title[\"head_tagged_en\"] = df_title[\"HEADLINE_ALERT_TEXT_y\"].apply(tagging_en)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: faster fast\n",
      "NN--ADJ error: brighter bright\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: leaner lean\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: better good\n",
      "NN--ADJ error: fresher fresh\n",
      "NN--ADJ error: fresher fresh\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: longer long\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: hardest hard\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: harder hard\n",
      "NN--ADJ error: smaller small\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: stronger strong\n",
      "NN--ADJ error: darker dark\n",
      "NN--ADJ error: costlier costly\n",
      "NN--ADJ error: costlier costly\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: lowest low\n",
      "NN--ADJ error: better good\n",
      "NN--ADJ error: louder loud\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: tougher tough\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: leaner lean\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: stronger strong\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: calmer calm\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: perkier perky\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: smartest smart\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: der d\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: slower slow\n",
      "NN--ADJ error: shorter short\n",
      "NN--ADJ error: shorter short\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: steadier steady\n",
      "NN--ADJ error: stricter strict\n",
      "NN--ADJ error: stricter strict\n",
      "NN--ADJ error: stricter strict\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: hardliner hardline\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: shorter short\n",
      "NN--ADJ error: shorter short\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: looser loose\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: brighter bright\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: cooler cool\n",
      "NN--ADJ error: cooler cool\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: faster fast\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: later late\n",
      "NN--ADJ error: larger large\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: latest late\n",
      "NN--ADJ error: shakier shaky\n",
      "NN--ADJ error: shakier shaky\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: calmer calm\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: stricter strict\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: biggest big\n",
      "NN--ADJ error: der d\n",
      "NN--ADJ error: tougher tough\n",
      "NN--ADJ error: bolder bold\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: stronger strong\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: slower slow\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: longer long\n",
      "NN--ADJ error: greater great\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: nearer near\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: kinder kind\n",
      "NN--ADJ error: better good\n",
      "NN--ADJ error: better good\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: calmer calm\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: highlander highland\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: cheapest cheap\n",
      "NN--ADJ error: clearer clear\n",
      "NN--ADJ error: clearer clear\n",
      "NN--ADJ error: clearer clear\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: hinder hind\n",
      "NN--ADJ error: hinder hind\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: cleaner clean\n",
      "NN--ADJ error: cleaner clean\n",
      "NN--ADJ error: gloomier gloomy\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: quicker quick\n",
      "NN--ADJ error: cheaper cheap\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: greater great\n",
      "NN--ADJ error: der d\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: latest late\n",
      "NN--ADJ error: cheaper cheap\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: cheaper cheap\n",
      "NN--ADJ error: tamer tame\n",
      "NN--ADJ error: tougher tough\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: cheaper cheap\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: faster fast\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: shorter short\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: slowest slow\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: nearer near\n",
      "NN--ADJ error: thinner thin\n",
      "NN--ADJ error: longer long\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: thinner thin\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: bolder bold\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: greater great\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: priciest pricy\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: easier easy\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: clearer clear\n",
      "NN--ADJ error: clearer clear\n",
      "NN--ADJ error: better good\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: broker broke\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: worst bad\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: buffer buff\n",
      "NN--ADJ error: sweeter sweet\n",
      "NN--ADJ error: higher high\n",
      "NN--ADJ error: gloomier gloomy\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: mester m\n",
      "NN--ADJ error: largest large\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: closer close\n",
      "NN--ADJ error: cooler cool\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: stabler stable\n",
      "NN--ADJ error: tighter tight\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: deeper deep\n",
      "NN--ADJ error: stronger strong\n",
      "NN--ADJ error: firmer firm\n",
      "NN--ADJ error: stricter strict\n",
      "NN--ADJ error: softer soft\n",
      "NN--ADJ error: lower low\n",
      "NN--ADJ error: cheaper cheap\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: wider wide\n",
      "NN--ADJ error: stronger strong\n",
      "NN--ADJ error: weaker weak\n",
      "NN--ADJ error: highlander highland\n",
      "NN--ADJ error: lighter light\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "data_preparation/retreive_news.py:237: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  df_title[\"head_tagged2_en\"] = df_title[\"head_tagged_en\"].apply(clean_tag_en)\n"
     ]
    }
   ],
   "source": [
    "run data_preparation/retreive_news.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/M2015eliu/.pyenv/versions/anaconda2-4.1.1/lib/python2.7/site-packages/ipykernel/__main__.py:16: FutureWarning: currently extract(expand=None) means expand=False (return Index/Series/DataFrame) but in a future version of pandas this will be changed to expand=True (return DataFrame)\n",
      "/home/M2015eliu/.pyenv/versions/anaconda2-4.1.1/lib/python2.7/site-packages/ipykernel/__main__.py:16: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[I] Finish referenece extraction for month 01\n",
      "[I] Finish referenece extraction for month 02\n",
      "[I] Finish referenece extraction for month 03\n",
      "[I] Finish referenece extraction for month 04\n",
      "[I] Finish referenece extraction for month 05\n",
      "[I] Finish referenece extraction for month 06\n",
      "[I] Finish referenece extraction for month 07\n"
     ]
    }
   ],
   "source": [
    "    for month in period:\n",
    "        # Load the file monthly\n",
    "        file_path = base_path + \"rna002_RTRS_\" + str(year) + \"_\" + month + \".csv\"\n",
    "        df = pd.read_csv(file_path, index_col=None, header=0)\n",
    "\n",
    "        # Filter out the Japanese articles\n",
    "        df_month_jp = df[df[\"LANGUAGE\"].isin([\"JA\"])\n",
    "                         & df[\"EVENT_TYPE\"].isin([\"STORY_TAKE_OVERWRITE\"])\n",
    "                         & df[\"TAKE_TEXT\"].str.contains(\"参照番号\")]\n",
    "\n",
    "        # Filter out the English articles\n",
    "        df_month_en = df[df[\"LANGUAGE\"].isin([\"EN\"])\n",
    "                         & df[\"EVENT_TYPE\"].isin([\"STORY_TAKE_OVERWRITE\"])]\n",
    "\n",
    "        # Extract the reference code for Japanese news [会有警告，为何？]\n",
    "        df_month_jp[\"reference_code\"] = df_month_jp[\"TAKE_TEXT\"].str.extract('参照番号\\\\[([\\\\w]+)\\\\]')\n",
    "\n",
    "        # Drop the row where reference code is NaN\n",
    "        df_month_jp = df_month_jp.dropna(subset=[\"reference_code\"])\n",
    "\n",
    "        # Find the English news basing on the Japanese news\n",
    "        df_month_merge = pd.merge(df_month_jp[column_need].reset_index(),\n",
    "                                  df_month_en[column_need[:-1]].reset_index(),\n",
    "                                  left_on='reference_code',\n",
    "                                  right_on='PNAC',\n",
    "                                  how=\"left\").dropna()\n",
    "        list_merge.append(df_month_merge)\n",
    "\n",
    "        # # Append the Data frame to the list\n",
    "        # list_jp.append(df_month_jp[column_need])\n",
    "        # list_en.append(df_month_en[column_need[:-1]])\n",
    "\n",
    "        print \"[I] Finish referenece extraction for month\", month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "20"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df_month_jp.iloc[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    ［香港　１日　ロイター］ - 民主派の大規模デモが続く香港で１日未明、政府本部庁舎を包囲しようとした数千人のデモ隊に警官隊が警棒で殴りつけたり、催涙スプレーを噴射したりした。 \n",
      "    政府本部庁舎のある香港島の金鐘（アドミラルティ）では、デモを主導する学生団体が抗議行動を強化しようと呼び掛け、呼応した群衆が集まっていた。 \n",
      "    警察当局によると、衝突に絡んで少なくとも１８人が逮捕された。 \n",
      "    多数の群衆は防護眼鏡などを身に着け、政府本部庁舎周辺地域からの撤退を拒否。「われわれは普通選挙を望んでいる」と叫び、警官隊と引き続き衝突している。 \n",
      "     \n",
      "    ※英文参照番号[nL3N0TK0AQ]（契約の内容によっては英文がご覧いただけない場合もあります） \n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(df_month_jp.iloc[1].TAKE_TEXT)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "香港デモ隊と警察が衝突、政府本部庁舎近く　少なくとも18人逮捕\n",
      "---\n",
      "    ［香港　１日　ロイター］ - 民主派の大規模デモが続く香港で１日未明、政府本部庁舎を包囲しようとした数千人のデモ隊に警官隊が警棒で殴りつけたり、催涙スプレーを噴射したりした。 \n",
      "    政府本部庁舎のある香港島の金鐘（アドミラルティ）では、デモを主導する学生団体が抗議行動を強化しようと呼び掛け、呼応した群衆が集まっていた。 \n",
      "    警察当局によると、衝突に絡んで少なくとも１８人が逮捕された。 \n",
      "    多数の群衆は防護眼鏡などを身に着け、政府本部庁舎周辺地域からの撤退を拒否。「われわれは普通選挙を望んでいる」と叫び、警官隊と引き続き衝突している。 \n",
      "     \n",
      "    ※英文参照番号[nL3N0TK0AQ]（契約の内容によっては英文がご覧いただけない場合もあります） \n",
      "\n",
      "UPDATE 4-Hong Kong protesters clash with police, government HQ closes\n",
      "---\n",
      " (Recasts with government headquarters closed, latest protests, \n",
      "paragraphs 1-13) \n",
      "    By Clare Baldwin and James Pomfret \n",
      "    HONG KONG, Dec 1 (Reuters) - Thousands of pro-democracy \n",
      "activists forced the temporary closure of the Hong Kong \n",
      "government's headquarters on Monday after they clashed with \n",
      "police outside, defying orders to retreat after more than two \n",
      "months of sustained protests. \n",
      "    Chaos erupted as commuters made their way to work, with \n",
      "hundreds of protesters surrounding Admiralty Centre, which \n",
      "houses offices and retail outlets, in a tense stand-off with \n",
      "police. The central government offices and the legislature were \n",
      "forced to close in the morning, as were scores of shops. \n",
      "    The latest flare-up, during which police charged protesters \n",
      "with batons and pepper spray, marked an escalation in the civil \n",
      "disobedience movement. It also underscored the frustration of \n",
      "protesters at Beijing's refusal to budge on electoral reforms \n",
      "and grant greater democracy to the former British colony.  \n",
      "    \"The atmosphere in Admiralty is very different now after the \n",
      "clashes last night,\" said Jessica Lam, 20, who returned to the \n",
      "protest site on Monday morning. \"It has become very tense, like \n",
      "back to the early days when the protest just started.\" \n",
      "    The democracy movement represents one of the biggest threats \n",
      "for China's Communist Party leadership since Beijing's bloody \n",
      "1989 crackdown on pro-democracy student protests in and around \n",
      "Tiananmen Square. \n",
      "    Hundreds of riot police scattered the crowds in several \n",
      "rounds of heated clashes overnight, forcing protesters back with \n",
      "pepper spray and batons as some tried to scramble over walls in \n",
      "a crush of bodies on a highway outside government headquarters.  \n",
      "    Scores of volunteer medics attended to numerous injured, \n",
      "some who lay unconscious and others with blood streaming from \n",
      "head gashes. Police said at least 40 arrests were made. \n",
      "    Hong Kong's security secretary Lai Tung-kwok defended the \n",
      "use of force: \"The police have to take resolute actions, they \n",
      "have no choice ... it is their duty to restore law and order.\" \n",
      "    As police tackled the running battles in Admiralty, tensions \n",
      "escalated across the harbour in the gritty working-class \n",
      "district of Mong Kok, which had been the scene of violent \n",
      "clashes in recent weeks before the clearance of a large protest \n",
      "encampment from a major road there last Wednesday.      \n",
      "    The protesters are demanding free elections for the city's \n",
      "next leader in 2017 rather than the vote between pre-screened \n",
      "candidates that Beijing has said it would allow. \n",
      "    The clashes came after student leaders called on activists \n",
      "to escalate their protests and surround government headquarters, \n",
      "galvanising supporters to make their way to the buildings in \n",
      "Admiralty, next to Hong Kong's central business district and \n",
      "some of the world's most expensive real estate. \n",
      "    Student representative Nathan Law urged protesters to \n",
      "continue the disobedience movement that began in late September, \n",
      "calling it a \"long journey\".   \n",
      "    Despite several waves of clampdowns, crowds of protesters, \n",
      "many in protective goggles and body armour, refused to leave the \n",
      "area and continued to press against police lines, chanting \"We \n",
      "want universal suffrage!\". They threw bottles, helmets and \n",
      "umbrellas at police as tensions simmered into mid-morning. \n",
      "    Scores of demonstrators held up umbrellas, which have become \n",
      "a symbol of the pro-democracy movement, to protect themselves \n",
      "from the pepper spray and batons. \n",
      "    The latest clashes highlight the challenges authorities face \n",
      "as a restive younger generation contests Beijing's grip on the \n",
      "financial hub and demands greater democracy. \n",
      "    The Hong Kong rallies drew more than 100,000 on to the \n",
      "streets at their peak. Numbers have since dwindled and public \n",
      "support for the movement has waned. \n",
      " \n",
      " (Additional reporting by Diana Chan, Clare Jim and Michelle \n",
      "Chen; Editing by Anne Marie Roantree and Paul Tait) \n",
      " ((annemarie.roantree@thomsonreuters.com; +852 97387151; Reuters \n",
      "Messaging: annemarie.roantree.thomsonreuters.com@reuters.net)) \n",
      " \n",
      "Keywords: HONGKONG CHINA/  \n",
      "     \n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(df_month_merge.iloc[1].HEADLINE_ALERT_TEXT_x)\n",
    "print(\"---\")\n",
    "print(df_month_merge.iloc[1].TAKE_TEXT_x)\n",
    "print(df_month_merge.iloc[1].HEADLINE_ALERT_TEXT_y)\n",
    "print(\"---\")\n",
    "print(df_month_merge.iloc[1].TAKE_TEXT_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "韓国の現代・起亜自、2014年世界販売台数は4％増見込む　03年以来の低成長\n",
      "---\n",
      "    ［ソウル　２日　ロイター］ - 韓国の自動車メーカー、現代自動車<005380.KS>とその系列の起亜自動車<005380.KS>は、２０１４年の世界での販売台数について、両社合計で前年比４％増の７８６万台との目標を設定した。内訳は現代が４９０万台で、起亜が２９６万台。規制当局に２日提出した文書で判明した。 \n",
      "    ４％という伸び率は、２．３％だった２００３年の以来の低水準。円安を背景に、日本のメーカーの輸出攻勢が強まることを織り込んだ。 \n",
      "    現代自の鄭夢九・会長は、従業員への新年のメッセージで「世界経済が低成長時代に入るなか、企業間の競争は激化している」と述べた。 \n",
      "    現代・起亜自の世界販売台数は、ウォン安で輸出競争力が高まった２０１０年には２４％増を記録した。２０１３年の販売台数は両社合わせて６％増の７５６万台で、２０１２年の８％増から伸びが鈍化した。 \n",
      "     \n",
      "    ※英文参照番号[nL3N0K90F2]（１３カ月以内であれば[ID:nL3N0K90F2]でご覧になれます。なお、契約の内容によっては、英文がご覧いただけない場合もあります） \n",
      "     \n",
      "\n",
      "UPDATE 2-Hyundai, Kia face fading growth as currency tides buoy Japan rivals\n",
      "---\n",
      "    * S. Korean firms see 2014 car sales at 7.86 mln, up 4 pct \n",
      "    * Quick weakening of yen favours exports by Japanese peers \n",
      "    * Shares skid with won at 5-year high vs dollar, yen \n",
      " \n",
      " (Add closing share prices, production figures, analyst \n",
      "comments) \n",
      "    By Hyunjoo Jin \n",
      "    SEOUL, Jan 2 (Reuters) - The man who led South Korea's auto \n",
      "industry on a tear through the last decade said Hyundai Motor Co \n",
      "<005380.KS> and Kia Motors Corp <000270.KS> expect what will be \n",
      "their lowest annual sales growth since 2003 as the weak yen \n",
      "fires up Japanese rivals. \n",
      "    In his annual New Year speech to staff on Thursday, \n",
      "75-year-old group chairman Chung Mong-koo said sales at Hyundai \n",
      "and its smaller affiliate Kia will likely grow just 4 percent in \n",
      "2014. Global competition is about to get tougher in an industry \n",
      "facing changing technology and an uncertain future, he warned.  \n",
      "    The Korean duo exports about two-thirds of its cars from \n",
      "South Korea, a much higher proportion than Japanese peers, \n",
      "making them more vulnerable to currency fluctuations. A weaker \n",
      "yen gives Japanese carmakers like Toyota Motor Corp <7203.T>, \n",
      "Honda Motor Co <7267.T> and Nissan Motor Co <7201.T> leeway to \n",
      "offer customers better deals in key export markets like the \n",
      "United States - the same advantage that helped Hyundai and Kia \n",
      "after the global financial crisis of 2008. \n",
      "    \"Competition among companies is intensifying, as the global \n",
      "economy has entered an era of low growth,\" Chung said in his \n",
      "traditional New Year address to some 1,000 employees at \n",
      "Hyundai's Seoul headquarters. \"Technological convergence leads \n",
      "to change in the industry, and adds to uncertainty.\" \n",
      "    On Thursday, the won <KRW=KFTC> hit its highest level for \n",
      "more than five years against both the dollar and the yen, \n",
      "prompting talk in Seoul of possible government intervention in \n",
      "currency markets to help exporters. Shares in Hyundai and Kia, \n",
      "ranked fifth in the world by combined global sales, ended down \n",
      "5.1 percent and 6.1 percent, respectively.   \n",
      "     \n",
      "<^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ \n",
      "    GRAPHIC: Growth Brake http://link.reuters.com/bah75v \n",
      "^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^>     \n",
      "    The yen's fall has been stoked by Japan's attempts to \n",
      "support its export industries and pull its economy out of a \n",
      "two-decade slump. A senior official in South Korea earlier this \n",
      "week expressed concerns that the country's own exporters could \n",
      "be hurt as a result: auto exports typically account for close to \n",
      "9 percent of South Korea's total. [ID:nL3N0K80IB] \n",
      "    Toyota is on course for record profits in the financial year \n",
      "ended March 2014, largely helped by the weaker yen. That trend \n",
      "also helped Mazda Motor Corp <7261.T> and Subaru maker Fuji \n",
      "Heavy Industries Ltd <7270.T> become two of the top three \n",
      "performers in the Nikkei stock average in 2013. \n",
      "     \n",
      "    BREAKNECK TO BLEAK \n",
      "    The year ahead promises to be testing for Chung, more than \n",
      "10 years into the job leading the company founded by his father. \n",
      "Hyundai has shaken up U.S. management amid a fall in market \n",
      "share there, while a new free trade deal with the European Union \n",
      "means German rivals are encroaching on its once-impregnable \n",
      "domestic stronghold.  \n",
      "    Hyundai and its affiliate Kia are planning for global sales \n",
      "of a combined 7.86 million vehicles this year, up from 7.56 \n",
      "million in 2013. To achieve that, Chung is counting on growth \n",
      "from revamped versions of key models like the Genesis and Sonata \n",
      "sedans, as well increased production capacity in China.  \n",
      "    The forecast increase for 2014 would be in line with the \n",
      "global auto market's projected sales growth of 4 percent, but is \n",
      "the lowest since 2003's 2.3 percent. It's also bleak by \n",
      "comparison with breakneck growth at one of the auto industry's \n",
      "biggest success stories during the global economic downturn at \n",
      "the end of the last decade.  \n",
      "    At their growth peak in 2010, Hyundai and Kia's annual sales \n",
      "climbed 24 percent as the weaker Korean won allowed them to \n",
      "offer stylish, feature-loaded new cars at affordable prices.   \n",
      "    \"The yen's weakness is not welcoming news to Hyundai \n",
      "investors. This may adversely affect its market share and \n",
      "sales,\" said Song Sun-jae, an analyst at Hana Daetoo Securities. \n",
      "    However, he said he still expects Hyundai and Kia to sell 8 \n",
      "million vehicles this year, higher than its target, and slightly \n",
      "outperform the global market. \"Hyundai and Kia have offered \n",
      "conservative sales targets and beat them in recent years. I \n",
      "expect this year would be no different,\" he said. \n",
      "    In regulatory filings, Hyundai said it was targeting sales \n",
      "of 4.9 million vehicles this year, while Kia's goal is to reach \n",
      "2.96 million. \n",
      "    The pair's 2013 sales did beat the original Hyundai-Kia \n",
      "target of 7.41 million as growth in China and Brazil offset a \n",
      "lackluster showing at home. In South Korea, auto makers like \n",
      "Germany's Volkswagen AG <VOWG_p.DE> have been boosted by a free \n",
      "trade pact between the country and the European Union.  \n",
      "    But at 6 percent, 2013's increase was below the previous \n",
      "year's 8 percent rise, leaving Hyundai-Kia already in their \n",
      "third straight year of declining growth. [ID:nL3N0K20OJ]         \n",
      "    Hyundai and Kia have increased overseas production over the \n",
      "years. They made 49 percent of their cars in Korea, according to \n",
      "industry data for 2012, the last year for which figures have \n",
      "been released. Of those, however, they still exported 67 \n",
      "percent. \n",
      "    By comparison, Honda makes about 20 percent of its cars in \n",
      "Japan, and of those it exports about 20 percent.  \n",
      " \n",
      "    CAPACITY CONSTRAINED \n",
      "    In recent years, Chung's focus has been more on managing \n",
      "quality and brand-building than new plants. But the export \n",
      "comparisons have left company watchers wondering whether Hyundai \n",
      "and Kia might announce plans to build new plants in the U.S. or \n",
      "other markets this year. \n",
      "    Fuelling those questions, analysts say capacity constraints, \n",
      "along with aging models and rising competition, may have led \n",
      "Hyundai to miss its 2013 sales target for the U.S. Automakers \n",
      "including Hyundai plan to announce U.S. sales on Friday. \n",
      "    The task of bolstering Hyundai in the U.S. in a year of \n",
      "crucial new launches now falls to sales head David Zuchowski, \n",
      "who recently replaced John Krafcik as U.S. chief. \n",
      "[ID:nL2N0K60V1] \n",
      "    Kia also plans to start production at its third plant in \n",
      "China early this year, while Hyundai has added capacity of \n",
      "150,000 vehicles at its third factory in Beijing. \n",
      " \n",
      "     \n",
      " \n",
      " (Additional reporting by Yoko Kubota in TOKYO and Se Young Lee \n",
      "in SEOUL; Editing by Kenneth Maxwell) \n",
      " ((hyunjoo.jin@thomsonreuters.com)(82-2-3704-5685)(Reuters \n",
      "Messaging: hyunjoo.jin.thomsonreuters.com@reuters.net)) \n",
      " \n",
      "Keywords: HYUNDAI SALES/  \n",
      "     \n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(df_merge.iloc[5].HEADLINE_ALERT_TEXT_x)\n",
    "print(\"---\")\n",
    "print(df_merge.iloc[5].TAKE_TEXT_x)\n",
    "print(df_merge.iloc[5].HEADLINE_ALERT_TEXT_y)\n",
    "print(\"---\")\n",
    "print(df_merge.iloc[5].TAKE_TEXT_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "第4四半期シンガポールＧＤＰ速報値、季節調整済み前期比年率‐2.7％　製造業が低迷\n",
      "---\n",
      "    ［シンガポール　２日　ロイター］ - シンガポール通産省が発表した２０１３年第 \n",
      "４・四半期国内総生産（ＧＤＰ）速報値は、季節調整済みの前期比年率２．７％減となっ \n",
      "た。製造業の低迷を背景に、２．２％増だった第３・四半期からマイナスに転じ、ロイタ \n",
      "ーが調査したエコノミスト予想中央値の１．６％減よりも深刻な落ち込みを示した。 \n",
      "    前年比は４．４％増で、第３・四半期の５．９％増から伸びが鈍化した。 \n",
      "    バンク・オブ・アメリカ・メリルリンチのアジア新興国経済責任者、Chua Hak Bin氏 \n",
      "は、「製造業が弱い。ようやく製造業が目に見える回復を示すと若干期待していたが、再 \n",
      "び失速したようだ。結局、相対的により安定しているサービス業が引き続き成長をけん引 \n",
      "している構図だ」と述べた。 \n",
      "    製造業は季節調整済みの前期比年率で４．０％減。前年比では３．５％増となったも \n",
      "のの、第３・四半期の５．３％増から鈍化した。 \n",
      "    サービス業は季節調整済みの前期比年率１．７％減、前年比５．５％増だった。 \n",
      "     \n",
      "    期間      Q4-2013   Q3-2013   Q2-2013   Q1-2013  Q4-2012 \n",
      "    Q/Q         -2.7      *2.2     *17.1     *2.2      3.3 \n",
      "    Y/Y         4.4       *5.9      *4.3      0.3      1.5 \n",
      " \n",
      "　　                 Q/Q       Y/Y \n",
      "   MANUFACTURING    -4.0       3.5 \n",
      "   CONSTRUCTION     -6.9       4.7 \n",
      "   SERVICES         -1.7       5.5 \n",
      " \n",
      "   市場予想中央値 \n",
      "    Q/Q         -1.6 \n",
      "    Y/Y         +4.7 \n",
      "                                                           \n",
      "　　 Q/Q ＝季節調整済み・年率 \n",
      "     * = 改定値 \n",
      "  \n",
      "　※英文参照番号[nL3N0KB0LT][nS7N0JA002]（１３カ月以内であれば[ID:nL3N0KB0LT][ID \n",
      ":nS7N0JA002]でご覧になれます。なお、契約の内容によっては、英文がご覧いただけない \n",
      "場合もあります）     \n",
      "     \n",
      "\n",
      "Singapore GDP contracts in Q4, hurt by sluggish manufacturing \n",
      "---\n",
      "    * Q4 GDP down s/adj annualised -2.7 pct q/q vs +2.2 pct in \n",
      "Q3 \n",
      "    * Q4 GDP +4.4 pct y/y vs +5.9 pct in Q3 \n",
      "    * Q4 manufacturing fell annualised s/adj 4.0 pct q/q \n",
      " \n",
      "    By Masayuki Kitano \n",
      "    SINGAPORE, Jan 2 (Reuters) - Singapore's economy contracted \n",
      "more than expected in the fourth quarter as manufacturing \n",
      "activity weakened, data showed on Thursday, casting some doubt \n",
      "on market expectations for a slight pick-up in growth over 2014. \n",
      "    According to advance estimates from Singapore's Ministry of \n",
      "Trade and Industry, GDP contracted an annualised and seasonally \n",
      "adjusted 2.7 percent in the final quarter of 2013 from the \n",
      "July-September period. \n",
      "    That reversed a 2.2 percent expansion in the third quarter \n",
      "and was weaker than the median 1.6 percent contraction forecast \n",
      "in a Reuters survey. \n",
      "    \"I think the weakness is in the manufacturing side. \n",
      "After...some hopes that the manufacturing recovery is finally \n",
      "showing up more visibly, it looks as if it's sputtering again,\" \n",
      "said Chua Hak Bin, head of emerging Asia economics for Bank of \n",
      "America Merrill Lynch. \n",
      "    \"Overall the story is that growth is still largely led by a \n",
      "more steady services sector,\" he added.      \n",
      "    The manufacturing sector contracted at an annualised rate of \n",
      "4.0 percent in the fourth quarter from the previous quarter, \n",
      "while year-on-year manufacturing growth moderated to 3.5 percent \n",
      "compared to 5.3 percent in the third quarter. \n",
      "    The services sector contracted an annualised 1.7 percent \n",
      "quarter-on-quarter, but grew 5.5 percent year-on-year.   \n",
      "    \"To put this in perspective it follows a very strong second \n",
      "quarter and a positive third quarter,\" said Credit Suisse \n",
      "economist Michael Wan, who is forecasting growth in 2014 will be \n",
      "4.0 percent, up on the 3.7 percent seen in 2013.  \n",
      "    Wan expects the economy to be supported over coming months \n",
      "by an improvement in external trade, although some sectors, \n",
      "especially retail and hospitality will feel the pressure of \n",
      "tighter government controls on hiring foreign labour.  \n",
      "    In the second quarter, Singapore's GDP had expanded by an \n",
      "annualised 17.1 percent quarter-on-quarter as manufacturing \n",
      "activity surged. \n",
      "    One question mark going forward is whether the manufacturing \n",
      "sector for Singapore is as leveraged to the global economy as it \n",
      "was in the past, said Chua at Bank of America Merrill Lynch. \n",
      "    \"It looks as if the restructuring and the stricter foreign \n",
      "worker policy is dampening growth,\" Chua said, adding that Bank \n",
      "of America Merrill Lynch's forecast is for full-year growth in \n",
      "2014 to slow to 3.2 percent, down from the 2013 growth of 3.7 \n",
      "percent. \n",
      "    At this stage, the Monetary Authority of Singapore looks \n",
      "likely to stick to its tight monetary policy stance at its next \n",
      "policy review in April, since inflation has been picking up, \n",
      "Chua added. \n",
      "    At its last policy-setting decision in October, Singapore's \n",
      "central bank maintained its stance of allowing a \"modest and \n",
      "gradual\" appreciation of the Singapore dollar and warned that \n",
      "core inflation was likely to accelerate in 2014.          \n",
      "    The Singapore government's forecast is for the economy to \n",
      "grow by 2-4 percent this year, while a central bank survey of \n",
      "forecasters in December showed that Singapore's economy is seen \n",
      "likely to expand at 3.9 percent in 2014 as exports recover. \n",
      " \n",
      " (Additional reporting by Rachel Armstrong; Editing by Eric \n",
      "Meijer) \n",
      " ((masayuki.kitano@thomsonreuters.com)(+65-6417-4682)(Reuters \n",
      "Messaging: masayuki.kitano.thomsonreuters.com@reuters.net)) \n",
      " \n",
      "Keywords: SINGAPORE ECONOMY/GDP \n",
      "     \n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(df_merge.iloc[6].HEADLINE_ALERT_TEXT_x)\n",
    "print(\"---\")\n",
    "print(df_merge.iloc[6].TAKE_TEXT_x)\n",
    "print(df_merge.iloc[6].HEADLINE_ALERT_TEXT_y)\n",
    "print(\"---\")\n",
    "print(df_merge.iloc[6].TAKE_TEXT_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "`Kia_NNP', \n",
      "`face_NN', \n",
      "`fading_VBG', \n",
      "`growth_NN', \n",
      "`as_IN', \n",
      "`currency_NN', \n",
      "`tides_NNS', \n",
      "`buoy_IN', \n",
      "`Japan_NNP', \n",
      "`rivals_NNS', \n"
     ]
    }
   ],
   "source": [
    "from nltk import *\n",
    "text = word_tokenize(\"Kia face fading growth as currency tides buoy Japan rivals\")\n",
    "for w,v in nltk.pos_tag(text):\n",
    "    print(\"`\"+w+\"_\"+v+\"', \")\n",
    "# print i [k+\"_\"+v for k,v in nltk.pos_tag(text)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "`Kia', \n",
      "`face', \n",
      "`fading', \n",
      "`growth', \n",
      "`as', \n",
      "`currency', \n",
      "`tides', \n",
      "`buoy', \n",
      "`Japan', \n",
      "`rivals', \n"
     ]
    }
   ],
   "source": [
    "for w in text:\n",
    "    print(\"`\"+w+\"', \")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "現代\n",
      "自動車\n",
      "と\n",
      "その\n",
      "系列\n",
      "の\n",
      "起亜\n",
      "自動車\n",
      "は\n",
      "、\n",
      "両社\n",
      "合計\n",
      "で\n",
      "前年\n",
      "比\n",
      "４\n",
      "％\n",
      "増\n",
      "の\n",
      "目標\n",
      "を\n",
      "設定\n",
      "し\n",
      "た\n",
      "\n"
     ]
    }
   ],
   "source": [
    "    import MeCab\n",
    "    text_jp = \"現代自動車とその系列の起亜自動車は、両社合計で前年比４％増の目標を設定した\"\n",
    "#     tagger = MeCab.Tagger(\"-Ochasen -d /usr/lib/mecab/dic/mecab-ipadic-neologd/\")\n",
    "    tagger = MeCab.Tagger()\n",
    "    node = tagger.parseToNode(text_jp)\n",
    "    line_tagged=[]\n",
    "    newLine = []\n",
    "    while node:\n",
    "        print node.surface\n",
    "        word_tagged = (node.surface, node.feature)\n",
    "        line_tagged.append(word_tagged)\n",
    "        list_feature = node.feature.split(',')\n",
    "        if '動詞' in list_feature[0] or '名詞' in list_feature[0] or '接頭詞' in list_feature[0]:\n",
    "            if '数' not in list_feature[1] and '接尾' not in list_feature[1]:\n",
    "                if '*' not in list_feature[6]:\n",
    "                    newLine.append(list_feature[6])\n",
    "        node = node.next\n",
    "\n",
    "    text_tagged_jp = ' '.join(newLine)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "``'', ``現代自動車'', ``と'', ``その'', ``系列'', ``の'', ``起亜自動車'', ``は'', ``、'', ``両社'', ``合計'', ``で'', ``前年比'', ``４'', ``％'', ``増'', ``の'', ``目標'', ``を'', ``設定'', ``し'', ``た'', ``'', \n"
     ]
    }
   ],
   "source": [
    "    import MeCab\n",
    "    text_jp = \"現代自動車とその系列の起亜自動車は、両社合計で前年比４％増の目標を設定した\"\n",
    "    tagger = MeCab.Tagger(\"-Ochasen -d /usr/lib/mecab/dic/mecab-ipadic-neologd/\")\n",
    "#     tagger = MeCab.Tagger()\n",
    "    node = tagger.parseToNode(text_jp)\n",
    "    line_tagged=[]\n",
    "    newLine = []\n",
    "    s = \"\"\n",
    "    while node:\n",
    "#         print node.surface, \":\", node.feature\n",
    "        s += \"``\"+node.surface + \"'', \"\n",
    "        word_tagged = (node.surface, node.feature)\n",
    "        line_tagged.append(word_tagged)\n",
    "        list_feature = node.feature.split(',')\n",
    "        if '動詞' in list_feature[0] or '名詞' in list_feature[0] or '接頭詞' in list_feature[0]:\n",
    "            if '数' not in list_feature[1] and '接尾' not in list_feature[1]:\n",
    "                if '*' not in list_feature[6]:\n",
    "                    newLine.append(list_feature[6])\n",
    "        node = node.next\n",
    "\n",
    "    text_tagged_jp = ' '.join(newLine)\n",
    "    print s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "``'', ``現代'', ``自動車'', ``と'', ``その'', ``系列'', ``の'', ``起亜'', ``自動車'', ``は'', ``、'', ``両社'', ``合計'', ``で'', ``前年'', ``比'', ``４'', ``％'', ``増'', ``の'', ``目標'', ``を'', ``設定'', ``し'', ``た'', ``'', \n"
     ]
    }
   ],
   "source": [
    "    import MeCab\n",
    "    text_jp = \"現代自動車とその系列の起亜自動車は、両社合計で前年比４％増の目標を設定した\"\n",
    "#     tagger = MeCab.Tagger(\"-Ochasen -d /usr/lib/mecab/dic/mecab-ipadic-neologd/\")\n",
    "    tagger = MeCab.Tagger()\n",
    "    node = tagger.parseToNode(text_jp)\n",
    "    line_tagged=[]\n",
    "    newLine = []\n",
    "    s = \"\"\n",
    "    while node:\n",
    "#         print node.surface, \":\", node.feature\n",
    "        s += \"``\"+node.surface + \"'', \"\n",
    "        word_tagged = (node.surface, node.feature)\n",
    "        line_tagged.append(word_tagged)\n",
    "        list_feature = node.feature.split(',')\n",
    "        if '動詞' in list_feature[0] or '名詞' in list_feature[0] or '接頭詞' in list_feature[0]:\n",
    "            if '数' not in list_feature[1] and '接尾' not in list_feature[1]:\n",
    "                if '*' not in list_feature[6]:\n",
    "                    newLine.append(list_feature[6])\n",
    "        node = node.next\n",
    "\n",
    "    text_tagged_jp = ' '.join(newLine)\n",
    "    print s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "``現代'', ``自動車'', ``系列'', ``起亜'', ``自動車'', ``世界'', ``販売'', ``台数'', ``両社'', ``合計'', ``前年'', ``目標'', ``設定'', ``する'', ``た'', \n"
     ]
    }
   ],
   "source": [
    "    import MeCab\n",
    "    text_jp = \"\t\t現代自動車$<005380.KS>$とその系列の起亜自動車$<005380.KS>$は、２０１４年の世界での販売台数について、両社合計で前年比４％増の７８６万台との目標を設定した\"\n",
    "#     tagger = MeCab.Tagger(\"-Ochasen -d /usr/lib/mecab/dic/mecab-ipadic-neologd/\")\n",
    "    tagger = MeCab.Tagger()\n",
    "    node = tagger.parseToNode(text_jp)\n",
    "    line_tagged=[]\n",
    "    newLine = []\n",
    "    s = \"\"\n",
    "    while node:\n",
    "#         print node.surface, \":\", node.feature\n",
    "#         s += \"``\"+node.surface + \"'', \"\n",
    "        word_tagged = (node.surface, node.feature)\n",
    "        line_tagged.append(word_tagged)\n",
    "        list_feature = node.feature.split(',')\n",
    "        if '動詞' in list_feature[0] or '名詞' in list_feature[0] or '接頭詞' in list_feature[0]:\n",
    "            if '数' not in list_feature[1] and '接尾' not in list_feature[1]:\n",
    "                if '*' not in list_feature[6]:\n",
    "                    newLine.append(list_feature[6])\n",
    "        node = node.next\n",
    "\n",
    "    text_tagged_jp = ' '.join(newLine)\n",
    "#     print text_tagged_jp.split()\n",
    "    for w in text_tagged_jp.split():\n",
    "        s += \"``\"+w + \"'', \"\n",
    "    print s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}