09_Basic_IE.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## [Basic] Information Extraction\n",
    "\n",
    "As an example task, consider the challenge of automating Amazon Retail's customer service email response. We should be able to find the following attributes or mark them as missing with high confidence:\n",
    "\n",
    "- Order Id\n",
    "- Dates (such as Shopping Date, Order Delivery) \n",
    "- Any `$` amounts \n",
    "\n",
    "Please note that I don't have any relation to Amazon other than shopping from there. \n",
    "\n",
    "Let's consider the following totally imagined complaint email from me to Jeff Bezos, the CEO of Amazon:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "complaint_email = \"\"\"Hello Jeff,\n",
    "\n",
    "I am Nirant, a loyal Amazon in first customer for months now. I am a huge fan of Kindle as well. \n",
    "I am stuck in a new city without a phone thanks to a sequence of problems - and are now compounded by Amazon's inhumane behaviour.\n",
    "\n",
    "The particular issues I am facing: My new phone bought from Amazon stopped working. What did I do? Requested a replacement on Jul 23\n",
    "- First Issue: The system did not allow a pick up on July 23 forcing a delay of more than a day to 24 July 8:00 - 11:00 AM\n",
    "- Second Issue: Despite requesting the customer service on chat THRICE, the pickup is delayed to July 24 8:00 - 11:00 AM\n",
    "- Third Issue: The pickup is rescheduled without any reason!\n",
    "\n",
    "Is this how you want Amazon to be world's most customer centric company?\n",
    "\n",
    "Here is how Amazon can help me:\n",
    "- Pick up the order as urgently as possible\n",
    "- Deliver the phone on a priority basis on Monday i.e. July 25 itself\n",
    "\n",
    "Here are the order numbers for reference: \n",
    "ORDER # 402-4870778-5154753 and ORDER # 404-8689779-9721113\n",
    "\n",
    "Here is my phone number: +91 7737887058\n",
    "\n",
    "I am stuck in a new city, where I don't know the language or directions without a working phone. I would really appreciate it if you could help in anyway. \n",
    "\n",
    "Regards,\n",
    "Nirant Kasliwal\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yikes, that is a lot of text. \n",
    "\n",
    "**The information to pull from this are (1) dates + times (2) phone number and (3) order numbers**. Let's figure out how to do that\n",
    "\n",
    "### Extract Date and Times\n",
    "\n",
    "If you are new to regex, consider reading the amazing [HOWTO on Python Regex](https://docs.python.org/3/howto/regex.html) and then coming back here. Let's warm up our regex muscles a bit: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['12', '11', '10']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = re.compile(r'\\d+')\n",
    "p.findall('12 drummers drumming, 11 pipers piping, 10 lords a-leaping')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['12', '11', '10']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = re.compile(r'\\d+')\n",
    "p.findall('12 drummers drumming, 11 pipers piping, 10 lords a-leaping')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TKX: Add compile and findall explanations here\n",
    "\n",
    "TKX: Add d+ explanations here"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 0 ns\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "date_pattern = r\"^(Jan|Feb|Mar|Apr|May|Jun|July|Aug|Sep|Oct|Nov|Dec)$\"\n",
    "p = re.compile(date_pattern)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p.findall(complaint_email)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "fastAI",
   "language": "python",
   "name": "fastai"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}