Why the 'AI revolution' is really a deep learning revolution.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "import numpy as np\n",
    "from collections import Counter\n",
    "import itertools\n",
    "\n",
    "def flatten(listOfLists):\n",
    "    \"Flatten one level of nesting\"\n",
    "    return itertools.chain.from_iterable(listOfLists)\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Why the 'AI revolution' is really a deep learning revolution\n",
    "\n",
    "## Introduction\n",
    "\n",
    "We often hear that there's been an explosion in AI research, that AI is advancing a breakneck speed.\n",
    "However, when we look at computer science publications we see no evidence of such changes in the AI research landscape.\n",
    "Infact many traditional fields such as logic programming, evolutionary algorithms, bayesian statistics appear to be on the decline.\n",
    "What we do see is a unprecidented explosion of publications in the field of deep learning.\n",
    "Since 2012 deep learning has gone from an unused term to a phrase that dominates the landscape.\n",
    "We estimate that more than one fith of AI or similar publications are now part of a deep learning gold rush.\n",
    "It is this explosion in deep learning that is fueling the latest media round of interest in AI.\n",
    "\n",
    "## Dataset\n",
    "\n",
    "arXiv is a free and open reposity for preprints of accademic papers that have been approved for publication after moderation. \n",
    "\n",
    "We take the titles of arXiv computer science publications tagged with any of the following catagories:\n",
    "\n",
    "* AI - Artificial Intelligence\n",
    "    * Covers all areas of AI except Vision, Robotics, Machine Learning, Multiagent Systems, and Computation and Language (Natural Language Processing), which have separate subject areas. In particular, includes Expert Systems, Theorem Proving (although this may overlap with Logic in Computer Science), Knowledge Representation, Planning, and Uncertainty in AI. \n",
    "* CL - Computation and Language (Computational Linguistics and Natural Language and Speech Processing)\n",
    "    * Covers natural language processing. \n",
    "* CV - Computer Vision and Pattern Recognition\n",
    "    * Covers image processing, computer vision, pattern recognition, and scene understanding.\n",
    "* LG - Machine Learning\n",
    "    * Covers machine learning and computational (PAC) learning. \n",
    "* NE - Neural and Evolutionary Computation\n",
    "    * Covers neural networks, connectionism, genetic algorithms, artificial life, adaptive behavior. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## uncomment and run the cell to download all of the data direct from arxiv (~1 hour)\n",
    "\n",
    "# ! python scrape.py > data.jsonlines\n",
    "# tags = [\"cs.AI\", \"cs.CL\", \"cs.CV\", \"cs.LG\", \"cs.NE\"]\n",
    "# data = pd.read_json('./data.jsonlines', lines=True)\n",
    "# data = data[data.catagories.apply(lambda records: any((item in records for item in tags)))]\n",
    "# data['year'] = data['created'].str.split('-').apply(lambda x: int(x[0]))\n",
    "# data = data[np.logical_and(data['year'] > 1996, data['year'] < 2018)]\n",
    "# data.to_json('filtered.jsonlines', lines=True, orient='records')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>year</th>\n",
       "      <th>catagories</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Towards a human eye behavior model by applying...</td>\n",
       "      <td>2008</td>\n",
       "      <td>[cs.HC, cs.NE]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Eye-Tracking Evolutionary Algorithm to minimiz...</td>\n",
       "      <td>2008</td>\n",
       "      <td>[cs.AI]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Node discovery in a networked organization</td>\n",
       "      <td>2008</td>\n",
       "      <td>[cs.AI]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Robustness and Regularization of Support Vecto...</td>\n",
       "      <td>2008</td>\n",
       "      <td>[cs.LG, cs.AI]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Multiagent Approach for the Representation of ...</td>\n",
       "      <td>2008</td>\n",
       "      <td>[cs.AI]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               title  year      catagories\n",
       "0  Towards a human eye behavior model by applying...  2008  [cs.HC, cs.NE]\n",
       "1  Eye-Tracking Evolutionary Algorithm to minimiz...  2008         [cs.AI]\n",
       "2         Node discovery in a networked organization  2008         [cs.AI]\n",
       "3  Robustness and Regularization of Support Vecto...  2008  [cs.LG, cs.AI]\n",
       "4  Multiagent Approach for the Representation of ...  2008         [cs.AI]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# otherwise load the prefilted data from the repo\n",
    "data = pd.read_json('./filtered.jsonlines', lines=True)[['title', 'year', 'catagories']]\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "43244"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing\n",
    "\n",
    "Below we create columns in the dataset to mark the presence of some key words."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data['is_deep'] = data['title'].str.contains('deep', case=False)\n",
    "data['is_adversarial'] = data['title'].str.contains('adversarial', case=False)\n",
    "data['is_reinforcement'] = data['title'].str.contains('reinforcement', case=False)\n",
    "data['is_convolutional'] = data['title'].str.contains('convolutional', case=False)\n",
    "data['is_ai'] = data['title'].str.contains(' ai ', case=False)\n",
    "data['is_artificial_intelligence'] = data['title'].str.contains('artificial intelligence', case=False)\n",
    "data['is_logic'] = data['title'].str.contains('logic', case=False)\n",
    "data['is_evolutionary'] = data['title'].str.contains('evolutionary', case=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Results\n",
    "\n",
    "### Word counts, in titles, per year\n",
    "\n",
    "Fuzzy logic and logical programming were staples of AI research through the late ninties and early 21st century.\n",
    "Below we can see a steady rate of publications containing the word `logic` in the title. \n",
    "There has been a gradual decline since 2002 as a share of the total number of publications."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEWCAYAAAB2X2wCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nLkmdMNAlqRP/H9GXlet1qazFAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a136cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = {}\n",
    "for year, group in data.groupby('year'):\n",
    "    counts[year] = {'freq': sum(group['is_logic'] > 0), 'total': len(group)}\n",
    "    \n",
    "hist = []\n",
    "years = sorted(counts.keys())    \n",
    "for year in years:\n",
    "    hist.append(100 * counts[year]['freq'] / counts[year]['total'])\n",
    "\n",
    "plt.xlabel('year')\n",
    "plt.title('Percentage of titles containing \"logic\" for each year')\n",
    "plt.bar(years, hist)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Similarly evolutionary algorithms peaked around 2004 with a gradual rise and fall either side."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEWCAYAAABv+EDhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVignADOzQv1/Lo1+OflPl4cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1056360b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = {}\n",
    "for year, group in data.groupby('year'):\n",
    "    counts[year] = {'freq': sum(group['is_evolutionary'] > 0), 'total': len(group)}\n",
    "    \n",
    "hist = []\n",
    "years = sorted(counts.keys())    \n",
    "for year in years:\n",
    "    hist.append(100 * counts[year]['freq'] / counts[year]['total'])\n",
    "\n",
    "plt.xlabel('year')\n",
    "plt.title('Percentage of titles containing \"evolutionary\" for each year')\n",
    "plt.bar(years, hist)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In stark contrast with the previous two examples titles containing the word deep have exploded."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEWCAYAAAB2X2wCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a20f2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = {}\n",
    "for year, group in data.groupby('year'):\n",
    "    counts[year] = {\n",
    "        'freq': sum(group['is_deep']),\n",
    "        'total': len(group)\n",
    "    }\n",
    "\n",
    "hist = []\n",
    "years = sorted(counts.keys())\n",
    "for year in years:\n",
    "    hist.append(100 * counts[year]['freq'] / counts[year]['total'])\n",
    "\n",
    "plt.xlabel('year')\n",
    "plt.title('Percentage of titles containing the \"deep\" for each year')\n",
    "\n",
    "plt.bar(years, hist)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When we examine several words indicative of deep learning research such as `deep`, `adversarial` and `convolutional` we see what appears to be exponential growth from around 2012."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAElCAYAAADjk4nIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzt3XmcHFW99/HPl4TNEIGQYUkI5AqI\nghcwjAHEBWULAQR9UOC6JIJGfOAqzwvU6PVCFLigz4MrXhEBAWVzi6IsIYK5yM4EA4TNBAwmJJAh\nAcLicsP9PX+cM6TSdPf0dPfMJKnv+/Xq11TXOVV16lTPr6tPdf9KEYGZmZXHeoPdADMzG1gO/GZm\nJePAb2ZWMg78ZmYl48BvZlYyDvxmZiXjwG8DSsmPJD0r6e4Glzlf0r/XKZ8m6Sfta+WaQ9KDkvZr\nd91WSbpE0pkDsa3CNkPSjgO5zXWVA38TJC2Q9FdJL0p6OgeyTQa7XUW5jQcMdjuqeAdwILBtRIyv\nLJQ0WdKtxXkRcUJEnJHL95O0aGCa2l6SxubgNbTRZSJi14iY1e66fVHtmLR5/bPycZ0maVp/bafB\ntuwnaVaeXmd/5OTA37zDI2ITYBzwNuDLfV1BXwLAOmR7YEFEvDTYDTErKwf+FkXEk8D1wFsAJG0q\n6SJJSyQ9KelMSUNy2WRJt0n6pqTlwLQ8/5OSHpb0gqSHJI3L80dJ+oWkbkl/lvSZnu3ms6OfSros\nL/egpM5c9mNgO+A3+VPJ5/P8n0l6StLzkm6RtGthfVtI+o2kFZLuye2+tVD+JkkzJS2X9KikD9Xq\nk9zua3Ld+ZI+mecfD1wI7JPb9ZWK5d4MnF8ofy7PvyS3Z1ju61G5/EVJo6psf29Jt0t6TtJ9xeGP\nfAwez332Z0kfrrEPQyR9SdJjue5sSWNy2dtzHz2f/769sNwsSWfk4/yCpBsljczFt+S/z+W27yNp\nB0k3S1om6RlJl0varLC+Vz+51TvmTdQdJ+mPuexnkq5WlaGbWsck21zStXkdd0naobBcw6+XWiR9\nTun/aLGk4yrKNpT0/yT9RelT9/mSNi6UHyZpTn4N3C5pt4p++qLS/9qzSp/YN+pr+9ZqEeFHHx/A\nAuCAPD0GeBA4Iz//FfADYBiwJXA38KlcNhlYCfwrMBTYGPgg8CTpU4OAHUlnxesBs4HTgA2ANwCP\nAwfndU0D/gZMBIYAZwN3VmtjYd5xwHBgQ+BbwJxC2VX58TpgF2AhcGsuG5affzy3exzwDLBrjf75\nL+A/gY2APYBuYP9CH9xap29fUw5cApyZp/cDFlWUTwN+kqdHA8tyv6xHGlZaBnTk/VgB7JzrblNn\nHz4HPADsnI/L7sAWwAjgWeCjuS+Ozc+3yMvNAh4D3piP7yzgnFw2FghgaGE7O+Y2bpjbeAvwrRqv\ntYaPeb26pNfTE8BngfWBDwD/6OnjPhyT5cD43A+XA1c183qpsc0JwNOkE6phwBW573bM5d8CrsnH\nYzjwG+DsXDYOWArslfd9Uu6bDQv9NJf0vzsCuK3Wvq+rj0FvwNr4yC+cF4Hn8j/Qf+Z/8q2AvwMb\nF+oeC/w+T08G/lKxrhnAZ6tsY68qdb8I/ChPTwN+VyjbBfhrRRsPqLMPm+V/pE3zP8d/kwNiLj+T\nVYH/aOAPFcv/ADi9ynrHAK8AwwvzzgYuKfRBfwb+LwA/rtLHk3IAeQ74X8VjVKMdjwJHVJn/UeDu\ninl3AJPz9Czgy4Wy/w3ckKfHUhH4q6z/SOCP1Y5jX455vbrAu0gnGyqU30rfA/+FhecTgUf6+nqp\n0w8Xk98w8/M35r7bkfRG/BKwQ6F8H+DPefr75BOxiuP57kI/nVDR9scabdu68CjjGHO7HBkRvyvO\nkPTPpDOoJZJ6Zq9HOvvpUZyGFCgfq7L+7UlDGsWP1kOAPxSeP1WYfhnYSNLQiFhZuTKl4aazSJ8w\nOoD/yUUjSW9aQ+u0c3tgr4q2DAV+XKXdo4DlEfFCYd4TQGeVuv1he+CDkg4vzFuf9Ob7kqSjgVOB\niyTdBpwSEY9UWU+t4zKKtD9FT5A+afSoPC41L/xL2hL4DvBO0pnreqRPELU0fMxr1c378GTkqJdV\nvi4bUWs/+/J6qWUU6RNvj2Kfd5A+mc4u/J+J9P/Rs/1Jkv61sMwGeZ09ivv7REXZOs+Bv70Wks74\nR9b4R4R01lK5zA5V6i0kncHs1GRbKrfzL8ARwAGkM55NSQFGpKGYlcC2wJ9y/TEVbfmviDiwge0u\nBkZIGl4I/tuRzjCbaXdfyxeSzvg/WXXhiBnAjDwefCbwQ1LQrbaeHUhDAkWLSYGlaDvghl7aBdXb\nfnaev1tELJN0JHBeA+tqxRJgtCQVgn+tNzrovc8r9eX1UssSVn8NbleYfgb4K2noqNrraiFwVkSc\nVWf9lete3GxD10a+uNtGEbEEuBE4V9LrJa2XL969u85iFwKnStpTyY6StiddG1gh6QuSNs4XG98i\n6W0NNudp0nWBHsNJb0rLSGdL/1Fo9yvAL4Fpkl4n6U3AxwrL/hZ4o6SPSlo/P96WL/xV9sFC4Hbg\nbEkb5Ytqx5PGgBtt97aSNqhTvoWkTWuU/wQ4XNLBuc82UvqK3raStpL0PqWLxH8nDde9UmM9FwJn\nSNopH5fdJG0BXJf74l8kDc2fIHYh9VFvukmftCqPy4ukC76jSdcW+tsdpP0+Ke/DEaSx+lp6OyaV\nGn691PFTYLKkXSS9Dji9pyAi/of0hv3N/IkJSaMlHZyr/BA4QdJe+dgNk3SopOGF9Z+YXxMjgC8B\nV9drjNLF8ll9aP8azYG//T5G+lj5EOmM+ueki4hVRcTPSEMwVwAvkC4Oj8jB+HDSxdE/k85yLiSd\nqTfibODL+VsNpwKXkT7SPpnbdmdF/ZPyup8ifSS/khQcyWfuBwHHkM6MngK+RrogWc2xpPHsxcB0\n0tjuzAbbfTPpYvlTkp6pLMzDMlcCj+d9G1VRvpD0yeZLpEC7kBRM18uPU3K7lgPvJo3BV/MNUvC5\nkXRB+CLSdYFlwGF5PcuAzwOHRcRr2lql7S+TjvVtue17A18hXYx8HriW9AbcryLiH6QLuseTrnl8\nhBSs/15jkbrHpMr6+/p6qbaO60kXcG8G5ue/RV/I8++UtAL4HelCPBHRBXyS9Mnp2VxvcsXyV5CO\n7eP50duP0caQLgKvE7T6MJ9ZIulrwNYRMWmw22L9T9JdwPkR8aPBbkt/k7QA+ETlNbpelplD+mba\nsn5r2ADyGb8Br37verf80Xg86Wxw+mC3y/qHpHdL2joP9UwCdqOx6xSlFBF7rCtBH3xx11YZThpC\nGUX6DvS5wK8HtUXWn3YmDWVtQrqoe1S+RmUl4KEeM7OS8VCPmVnJOPC3kQYhVW0z1LdUv06FW4Va\nyH4qaTulvDdDeq/dPLUhq2b+CuwtSvl4zm1X29rNr9O+ceDP8j/y2By8Jw92e/pTNJm+t6dv8uOS\n9rfs1dTFC/L0Aklj+2M7A6nyTSIi/hIRm+Sv7A5kOyL/ndXoGz8whfRV4tdHxCn91TYbWA78a7j8\nLZu2HCetYWmg17T2WFXbAw9FExcDy3581+T9d+CvQ72nzH2rpHvzx+CrSdkoe8oelnRY4fnQvI6e\nlMv1UgfPknSWUi6Zl4E3qEY64QbauEDp17/3Ay/ldhTT946XdEduxxJJ56nxX2jW6rdPKqVjXq6U\nnnlUoSwknShpHjCvD+vcWNK5kp5QSod8q3IaXqVf4z6Y92GWCr8Qzft6qqT783JXK6fgbeAY1Vxv\nRdtWG+JT4WYxqpIiWxU3ZFGNNNa5rLf0ylO1KnX0Q5Le32if9iZ/qpsEfD63/QCldMjfUkqVvDhP\nb1jc7/x6ewqo+psAScflvn9W0gylX6r3lH1b0kKl9OCzJb2zUFYzVXZ2gKR5eb3fk1Yl8imsY2tJ\nLyv9Crtn3p5Kqc/Xb7F90yT9XNJPlH5UNrnPnT5QBjtL3Jr8oE7KXFaltv0/pCRgR5EyXPZkkTwN\nuLywrkNZlb2wZurgXD4L+AuwK+krt5tSI51wvTbm8gXAHNIvDzcuzOvJ4rgnsHfezljgYeDkwvKv\npsJtsM/eSxoaGJfb9F3glor1zSSlw62bIbNivd/L/TKalIzr7Xn9byRlajwwH4fPk36puUFhX+8m\nfU11RN6/Exo4Ro2st6cPL6GQ2ZKKDKJUZEqlIksn9dNYT6N+KuYP5n1bj5QV8yVgm1w2mTqZUBvs\n98p9+yrpV99b5tfb7axKSb4fKedTz690X3N8SdlH5wNvzq+5LwO3F8o/Qkp/PZT06+ingI1yWdVU\n2YXX1W9JWWe3y304ocY+XQd8uvD8m8B329C+aaQYcGQ+Hg2/vgf6MegNWJseFFLmklLbLmb11La3\nsyrw70hKwfC6/Pxy4LQ8XTN1cJ6eBXy1UNaXdMLV0voeV1FnATVSNgMnA9MLz/sa+C8Cvl54vkn+\nZxhbWN97+9jv65GScu1epezfgZ9W1H0S2K+wrx8plH+d9AvV3o5RI+ttOfDTexrradRJxVylP+aQ\n00nTP4H/MWBi4fnBpDuq9ez3P8iBsMb6rgeOr+jXl4Hta9R/tue4UyNVduF19Y7C858CU2vUPRq4\nLU8PIQXv8W1o3zQKJzlr8sNDPXVI2lLSVUp30lpBSgDWczelaqltX00dGxHzSWeXhyslmXofKT8I\nrEod/FzPg3Qv2mJOn4WFdb1EerGeQEr5fK1SIrXe2viadVXZxzdK+q3SnblWkJK3VS7fF6ulLY6I\nF0mfZoppi/uaAngk6Wy41zTJkRJ4LaSBNMm9HKNG1tsOtdJY12t/T3plJH1Mq+409RzpxiWtHL9G\n2ltMkVyZ0rg7Iv5WZ/ntgW8X2rucdPY+GkDSKXmY5flcvimr9qdeBlFoPB32r4FdJL2B9Inu+Yi4\nuw3tg+bSWw84B/76iilzX0/6mNczbvhqattC/e0qlr+SlLDsCNIFsvl5fk/q4M0Kj2ERcU5h2dUu\npkXEjEhpbrcBHiFlIOytjVXXVeH7eX075eW/VGX5vlgtbbFSJswtWD0tc18vFD5DGu6olr66cnsi\nBYhG00DXOkZ9We9LpIynPbauKK+3v6+msS7MayiNdR57/iEpwd4WEbEZKY10K8evN5VpqStTGvd2\nbBeS7khXfO1vHBG35/HyLwAfAjbP+/M8q/anVgrzPslvTD8FPky6sU7xPgGttA/6/toeFA789dVL\nmXsHaTzzM0oXBT/Aa1PbXkXKUvhpVp1JQp3UwdUaofrphFtN6zucdP3gxfwp4tN9XL7SFcDHJe2R\nL/r9B3BXRCxodoX5bPti4BtKF0KHKN2vdkPSP/ChkvbPF+dOIfXR7Q2uvtYx6st65wATJY2QtDVp\nuKyoMkV2cd9aSWM9jBRougEkfZx87+d+dCUp62uH0r2ETyO9nht1PvBF5fs9K92j+oO5bDjpf6ob\nGCrpNOD1hWVrpcpuxmWkobD3VbS/lfatNRz466uZMjdWpbadTBrnO5qKlLqRcp/cQboQeXVhfr3U\nwdXUSyfcalrfU0k3aXmBdPZYNy95byLiJtL4+C9In4p2IKXnbdWppAt795D64GvAehHxKOlTzndJ\nnwwOBw7Px6eR9tY6Rn1Z74+B+0hj+Tfy2j6sTJFdqak01hHxECmn0h2kN5d/pv9TB58JdAH3k47H\nvfSe0vhVETGddOyuykOLc4FDcvEM0hj7n0hDSH9j9aGTqqmym9mJiLiNdG+Ee4snJS22b63hXD1m\nVkqSbgauiIgLB7stA82B38xKR+lOdjOBMRUX1kvBQz1mViqSLiXdsevkMgZ98Bm/mVnp+IzfzKxk\n1sgkQiNHjoyxY8cOdjPMzNYas2fPfiYiOhqpu0YG/rFjx9LV1TXYzTAzW2tIeqL3WomHeszMSqbX\nwC9pjKTf5/wUD0r6bJ4/QtJMpTSoMyVtXmP5SbnOPEmT2r0DZmbWN42c8a8ETomIN5PS954oaRdg\nKnBTROwE3JSfr0bSCOB0YC9SOoPTa71BmJnZwOg18EfEkoi4N0+/QMpmOJqUcuDSXO1SUjrgSgcD\nMyNieUQ8S/rBxIR2NNzMzJrTpzF+pfufvhW4C9gq5znpyXeyZZVFRrN6LotFtD+trZmZ9UHDgV/S\nJqTEWydHxIpGF6syr+ovxiRNkdQlqau7u7vRZpmZWR81FPhzWtpfkG5T15P98WlJ2+TybYClVRZd\nRMph3mNbVs/d/aqIuCAiOiOis6Ojoa+implZExr5Vo9I6U8fjohvFIquId2Imfz311UWnwEcJGnz\nfFH3oDzPzMwGSSNn/PuS7lLz3nyLtzmSJgLnAAdKmke6fdk5AJI6JV0IEBHLgTNIOdTvId1Hdnk/\n7IeZmTVojUzS1tnZGf7lrpmt7cZOvbZP9Recc2jT25I0OyI6G6nrX+6amZWMA7+ZWck48JuZlYwD\nv5lZyTjwm5mVjAO/mVnJOPCbmZWMA7+ZWck48JuZlYwDv5lZyTjwm5mVjAO/mVnJOPCbmZWMA7+Z\nWck48JuZlYwDv5lZyTjwm5mVjAO/mVnJDO2tgqSLgcOApRHxljzvamDnXGUz4LmI2KPKsguAF4BX\ngJWN3hbMzMz6T6+BH7gEOA+4rGdGRBzdMy3pXOD5Osu/JyKeabaBZmbWXr0G/oi4RdLYamWSBHwI\neG97m2VmZv2l1TH+dwJPR8S8GuUB3ChptqQp9VYkaYqkLkld3d3dLTbLzMxqaTXwHwtcWad834gY\nBxwCnCjpXbUqRsQFEdEZEZ0dHR0tNsvMzGppOvBLGgp8ALi6Vp2IWJz/LgWmA+Ob3Z6ZmbVHK2f8\nBwCPRMSiaoWShkka3jMNHATMbWF7ZmbWBr0GfklXAncAO0taJOn4XHQMFcM8kkZJui4/3Qq4VdJ9\nwN3AtRFxQ/uabmZmzWjkWz3H1pg/ucq8xcDEPP04sHuL7TMzszbzL3fNzErGgd/MrGQc+M3MSsaB\n38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzErGgd/M\nrGQc+M3MSsaB38ysZBz4zcxKppFbL14saamkuYV50yQ9KWlOfkyssewESY9Kmi9pajsbbmZmzWnk\njP8SYEKV+d+MiD3y47rKQklDgO8BhwC7AMdK2qWVxpqZWet6DfwRcQuwvIl1jwfmR8TjEfEP4Crg\niCbWY2ZmbdTKGP9Jku7PQ0GbVykfDSwsPF+U51UlaYqkLkld3d3dLTTLzMzqaTbwfx/YAdgDWAKc\nW6WOqsyLWiuMiAsiojMiOjs6OppslpmZ9aapwB8RT0fEKxHxP8APScM6lRYBYwrPtwUWN7M9MzNr\nn6YCv6RtCk/fD8ytUu0eYCdJ/yRpA+AY4JpmtmdmZu0ztLcKkq4E9gNGSloEnA7sJ2kP0tDNAuBT\nue4o4MKImBgRKyWdBMwAhgAXR8SD/bIXZmb9aOzUa/tUf8E5h/ZTS9qj18AfEcdWmX1RjbqLgYmF\n59cBr/mqp5mZDR7/ctfMrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB\n38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzEqm18Av\n6WJJSyXNLcz7v5IekXS/pOmSNqux7AJJD0iaI6mrnQ03M7PmNHLGfwkwoWLeTOAtEbEb8Cfgi3WW\nf09E7BERnc010czM2qnXwB8RtwDLK+bdGBEr89M7gW37oW1mZtYP2jHGfxxwfY2yAG6UNFvSlHor\nkTRFUpekru7u7jY0y8zMqmkp8Ev6N2AlcHmNKvtGxDjgEOBESe+qta6IuCAiOiOis6Ojo5VmmZlZ\nHU0HfkmTgMOAD0dEVKsTEYvz36XAdGB8s9szM7P2aCrwS5oAfAF4X0S8XKPOMEnDe6aBg4C51eqa\nmdnAaeTrnFcCdwA7S1ok6XjgPGA4MDN/VfP8XHeUpOvyolsBt0q6D7gbuDYibuiXvTAzs4YN7a1C\nRBxbZfZFNeouBibm6ceB3VtqnZmZtZ1/uWtmVjIO/GZmJePAb2ZWMr2O8ZuZrQvGTr22T/UXnHNo\nP7Vk8PmM38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjN\nzErGgd/MrGQc+M3MSsaB38ysZBoK/JIulrRU0tzCvBGSZkqal/9uXmPZSbnOvHyDdjMzG0SNnvFf\nAkyomDcVuCkidgJuys9XI2kEcDqwFzAeOL3WG4SZmQ2MhgJ/RNwCLK+YfQRwaZ6+FDiyyqIHAzMj\nYnlEPAvM5LVvIGZmNoBaGePfKiKWAOS/W1apMxpYWHi+KM97DUlTJHVJ6uru7m6hWWZmVk9/X9xV\nlXlRrWJEXBARnRHR2dHR0c/NMjMrr1YC/9OStgHIf5dWqbMIGFN4vi2wuIVtmplZi1oJ/NcAPd/S\nmQT8ukqdGcBBkjbPF3UPyvPMzGyQNPp1ziuBO4CdJS2SdDxwDnCgpHnAgfk5kjolXQgQEcuBM4B7\n8uOreZ6ZmQ2SoY1UiohjaxTtX6VuF/CJwvOLgYubap2ZmbWdf7lrZlYyDvxmZiXjwG9mVjIO/GZm\nJePAb2ZWMg78ZmYl48BvZlYyDvxmZiXjwG9mVjIO/GZmJePAb2ZWMg78ZmYl48BvZlYyDvxmZiXj\nwG9mVjIO/GZmJePAb2ZWMk0Hfkk7S5pTeKyQdHJFnf0kPV+oc1rrTTYzs1Y0dOvFaiLiUWAPAElD\ngCeB6VWq/iEiDmt2O2Zm1l7tGurZH3gsIp5o0/rMzKyftCvwHwNcWaNsH0n3Sbpe0q5t2p6ZmTWp\n5cAvaQPgfcDPqhTfC2wfEbsD3wV+VWc9UyR1Serq7u5utVlmZlZDO874DwHujYinKwsiYkVEvJin\nrwPWlzSy2koi4oKI6IyIzo6OjjY0y8zMqmlH4D+WGsM8kraWpDw9Pm9vWRu2aWZmTWr6Wz0Akl4H\nHAh8qjDvBICIOB84Cvi0pJXAX4FjIiJa2aaZmbWmpcAfES8DW1TMO78wfR5wXivbMDOz9vIvd83M\nSsaB38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzErG\ngd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGRaDvySFkh6QNIcSV1VyiXpO5Lm\nS7pf0rhWt2lmZs1r6Z67Be+JiGdqlB0C7JQfewHfz3/NzPps7NRrG6674JxD+7Ela6+BGOo5Args\nkjuBzSRtMwDbNTOzKtoR+AO4UdJsSVOqlI8GFhaeL8rzViNpiqQuSV3d3d1taJaZmVXTjsC/b0SM\nIw3pnCjpXRXlqrJMvGZGxAUR0RkRnR0dHW1olpmZVdNy4I+IxfnvUmA6ML6iyiJgTOH5tsDiVrdr\nZmbNaSnwSxomaXjPNHAQMLei2jXAx/K3e/YGno+IJa1s18zMmtfqt3q2AqZL6lnXFRFxg6QTACLi\nfOA6YCIwH3gZ+HiL2zQzsxa0FPgj4nFg9yrzzy9MB3BiK9sxM7P28S93zcxKxoHfzKxkHPjNzErG\ngd/MrGQc+M3MSsaB38ysZBz4zcxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4zcxKxoHf\nzKxkHPjNzErGgd/MrGQc+M3MSqbpwC9pjKTfS3pY0oOSPlulzn6Snpc0Jz9Oa625ZmbWqlZuvbgS\nOCUi7s03XJ8taWZEPFRR7w8RcVgL2zEzszZq+ow/IpZExL15+gXgYWB0uxpmZmb9oy1j/JLGAm8F\n7qpSvI+k+yRdL2nXOuuYIqlLUld3d3c7mmVmZlW0HPglbQL8Ajg5IlZUFN8LbB8RuwPfBX5Vaz0R\ncUFEdEZEZ0dHR6vNMjOzGloK/JLWJwX9yyPil5XlEbEiIl7M09cB60sa2co2zcysNa18q0fARcDD\nEfGNGnW2zvWQND5vb1mz2zQzs9a18q2efYGPAg9ImpPnfQnYDiAizgeOAj4taSXwV+CYiIgWtmlm\nZi1qOvBHxK2AeqlzHnBes9sws3XT2KnX9qn+gnMO7aeWlJN/uWtmVjIO/GZmJePAb2ZWMg78ZmYl\n48BvZlYyDvxmZiXjwG9mVjIO/GZmJePAb2ZWMg78ZmYl48BvZlYyrSRps8x5R8xsbeIzfjOzkvEZ\nv5k1xZ90114+4zczKxmf8ZutA3z2bX3hwF9CDhJm5eahHjOzkmnpjF/SBODbwBDgwog4p6J8Q+Ay\nYE/STdaPjogFrWzTbE3Xl09UxU9T/iRmA6XpwC9pCPA94EBgEXCPpGsi4qFCteOBZyNiR0nHAF8D\njm6lwWbWXs2+Udnaq5WhnvHA/Ih4PCL+AVwFHFFR5wjg0jz9c2B/SXVv0G5mZv1LEdHcgtJRwISI\n+ER+/lFgr4g4qVBnbq6zKD9/LNd5psr6pgBT8tOdgUebati6YyTwmn6y1biPeuc+6t260kfbR0RH\nIxVbGeOvduZe+S7SSJ00M+IC4IIW2rNOkdQVEZ2D3Y41mfuod+6j3pWxj1oZ6lkEjCk83xZYXKuO\npKHApsDyFrZpZmYtaiXw3wPsJOmfJG0AHANcU1HnGmBSnj4KuDmaHVsyM7O2aHqoJyJWSjoJmEH6\nOufFEfGgpK8CXRFxDXAR8GNJ80ln+se0o9El4WGv3rmPeuc+6l3p+qjpi7tmZrZ28i93zcxKxoHf\nzKxkHPgHiKQxkn4v6WFJD0r6bJ4/QtJMSfPy383zfEn6jqT5ku6XNK6wrkm5/jxJk2ptc23T5j56\nRdKc/Kj80sFaq4k+epOkOyT9XdKpFeuaIOnR3H9TB2N/+kOb+2iBpAfy66hrMPanX0SEHwPwALYB\nxuXp4cCfgF2ArwNT8/ypwNfy9ETgetJvIfYG7srzRwCP57+b5+nNB3v/1qQ+ymUvDvb+rCF9tCXw\nNuAs4NTCeoYAjwFvADYA7gN2Gez9W5P6KJctAEYO9j61++Ez/gESEUsi4t48/QLwMDCa1dNaXAoc\nmaePAC6L5E5gM0nbAAcDMyNqEhUVAAADR0lEQVRieUQ8C8wEJgzgrvSbNvbROquvfRQRSyPiHuC/\nK1bVSMqVtVIb+2id5cA/CCSNBd4K3AVsFRFLIL1gSWcfkF6oCwuLLcrzas1fp7TYRwAbSeqSdKek\nI1kHNdhHtfh11HsfQco0cKOk2TmtzDrBN2IZYJI2AX4BnBwRK+rkrKuV7qLhNBhrqzb0EcB2EbFY\n0huAmyU9EBGP9UNzB0Uf+qjmKqrMK+vrqJ598+toS2CmpEci4pa2NnQQ+Ix/AElan/RCvDwifpln\nP90zPJH/Ls3za6XEaCRVxlqrTX1ERPT8fRyYRTrrWyf0sY9q8euo9z4qvo6WAtNJQ2RrPQf+AaJ0\nunER8HBEfKNQVExrMQn4dWH+x/I3V/YGns8fT2cAB0naPH8r4aA8b63Xrj7KfbNhXudIYF+geJ+I\ntVYTfVRLIylX1krt6iNJwyQN75km/a/NbX+LB8FgX10uywN4B+mj9P3AnPyYCGwB3ATMy39H5Poi\n3ejmMeABoLOwruOA+fnx8cHetzWtj4C35+f35b/HD/a+DWIfbU06u18BPJenX5/LJpK+8fIY8G+D\nvW9rWh+RvvF0X348uC71kVM2mJmVjId6zMxKxoHfzKxkHPjNzErGgd/MrGQc+M3MSsaB38ysZBz4\nzfqBpCGD3QazWhz4rfQkndGTsz0/P0vSZyR9TtI9Odf/Vwrlv8pJux4sJu6S9KKkr0q6C9hngHfD\nrGEO/Gbp5/2TACStR0pf8DSwEyk3yx7AnpLelesfFxF7Ap3AZyRtkecPA+ZGxF4RcetA7oBZXzg7\np5VeRCyQtEzSW4GtgD+SbsxxUJ4G2IT0RnALKdi/P88fk+cvA14hJQYzW6M58JslFwKTSXlbLgb2\nB86OiB8UK0naDzgA2CciXpY0C9goF/8tIl4ZqAabNctDPWbJdNKdzN5GynY6Azgu53RH0uick31T\n4Nkc9N9EuuWj2VrFZ/xmQET8Q9LvgefyWfuNkt4M3JFv4PEi8BHgBuAESfcDjwJ3DlabzZrl7Jxm\nvHpR917ggxExb7DbY9afPNRjpSdpF9K9DW5y0Lcy8Bm/mVnJ+IzfzKxkHPjNzErGgd/MrGQc+M3M\nSsaB38ysZP4/Yw11DqkxJVUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x108caed30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = {}\n",
    "for year, group in data.groupby('year'):\n",
    "    counts[year] = {\n",
    "        'freq': sum(group['is_deep'] | group['is_adversarial'] | group['is_convolutional']),\n",
    "        'total': len(group)\n",
    "    }\n",
    "\n",
    "hist = []\n",
    "years = sorted(counts.keys())\n",
    "for year in years:\n",
    "    hist.append(100 * counts[year]['freq'] / counts[year]['total'])\n",
    "    \n",
    "plt.bar(years, hist)\n",
    "plt.xlabel('year')\n",
    "plt.title('Percentage of titles containing the \"deep\",\\n\"adversarial\" or \"convolutional\"  for each year')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As a side note observe that the terms such as `artificial_intelligence` or `ai` are not commonly used in titles.\n",
    "138 out of 41000 titles contain one of these terms (fewer than 0.5%)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(138, 43244)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data[data['is_artificial_intelligence'] | data['is_ai']]), len(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Which words have changed?\n",
    "\n",
    "Next we examine which words have seen the biggest increase or decreate in usage pre and post 2012"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "base_cv = CountVectorizer(analyzer='word', stop_words='english', ngram_range=[1, 2])\n",
    "base_cv = base_cv.fit(data['title'])\n",
    "\n",
    "cv = CountVectorizer(analyzer='word', stop_words='english', ngram_range=[1,2], vocabulary=base_cv.vocabulary_.keys())\n",
    "arr = cv.fit_transform(data[data['year'] < 2012]['title'])\n",
    "old_counts = np.array(np.sum(arr, axis=0)).flatten() / arr.shape[0]\n",
    "\n",
    "cv = CountVectorizer(analyzer='word', stop_words='english', ngram_range=[1, 2], vocabulary=base_cv.vocabulary_.keys())\n",
    "arr = cv.fit_transform(data[data['year'] > 2012]['title'])\n",
    "new_counts = np.array(np.sum(arr, axis=0)).flatten()  / arr.shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "diff = 100 * (new_counts - old_counts)\n",
    "keys = list(base_cv.vocabulary_.keys())\n",
    "arg_diff = np.argsort(diff)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pre - 2012"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>old freq (%)</th>\n",
       "      <th>new freq (%)</th>\n",
       "      <th>ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>logic</td>\n",
       "      <td>3.200746</td>\n",
       "      <td>0.809009</td>\n",
       "      <td>0.252756</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>systems</td>\n",
       "      <td>3.060907</td>\n",
       "      <td>1.478738</td>\n",
       "      <td>0.483104</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>programming</td>\n",
       "      <td>2.284027</td>\n",
       "      <td>0.820862</td>\n",
       "      <td>0.359393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>algorithm</td>\n",
       "      <td>4.195152</td>\n",
       "      <td>2.735220</td>\n",
       "      <td>0.651995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>information</td>\n",
       "      <td>3.076445</td>\n",
       "      <td>1.686176</td>\n",
       "      <td>0.548092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>constraint</td>\n",
       "      <td>1.755749</td>\n",
       "      <td>0.435620</td>\n",
       "      <td>0.248111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>algorithms</td>\n",
       "      <td>3.091983</td>\n",
       "      <td>1.834346</td>\n",
       "      <td>0.593259</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>theory</td>\n",
       "      <td>2.066501</td>\n",
       "      <td>0.811972</td>\n",
       "      <td>0.392921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>language</td>\n",
       "      <td>3.262896</td>\n",
       "      <td>2.142540</td>\n",
       "      <td>0.656637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>complexity</td>\n",
       "      <td>1.709136</td>\n",
       "      <td>0.663802</td>\n",
       "      <td>0.388385</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          word  old freq (%)  new freq (%)     ratio\n",
       "0        logic      3.200746      0.809009  0.252756\n",
       "1      systems      3.060907      1.478738  0.483104\n",
       "2  programming      2.284027      0.820862  0.359393\n",
       "3    algorithm      4.195152      2.735220  0.651995\n",
       "4  information      3.076445      1.686176  0.548092\n",
       "5   constraint      1.755749      0.435620  0.248111\n",
       "6   algorithms      3.091983      1.834346  0.593259\n",
       "7       theory      2.066501      0.811972  0.392921\n",
       "8     language      3.262896      2.142540  0.656637\n",
       "9   complexity      1.709136      0.663802  0.388385"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(\n",
    "    [(keys[i], 100*old_counts[i], 100*new_counts[i], new_counts[i]/old_counts[i]) for i in arg_diff[:10]], \n",
    "    columns = ['word', 'old freq (%)', 'new freq (%)', 'ratio']\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Post - 2012"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/dstaff/anaconda3/lib/python3.6/site-packages/ipykernel_launcher.py:2: RuntimeWarning: divide by zero encountered in double_scalars\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>old freq (%)</th>\n",
       "      <th>new freq (%)</th>\n",
       "      <th>ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>deep learning</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.169210</td>\n",
       "      <td>inf</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>detection</td>\n",
       "      <td>2.377253</td>\n",
       "      <td>4.637724</td>\n",
       "      <td>1.950875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>convolutional neural</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.290710</td>\n",
       "      <td>inf</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>image</td>\n",
       "      <td>2.843381</td>\n",
       "      <td>5.746036</td>\n",
       "      <td>2.020846</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>neural networks</td>\n",
       "      <td>1.258546</td>\n",
       "      <td>4.868869</td>\n",
       "      <td>3.868647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>convolutional</td>\n",
       "      <td>0.015538</td>\n",
       "      <td>3.834642</td>\n",
       "      <td>246.797570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>neural</td>\n",
       "      <td>3.107520</td>\n",
       "      <td>9.313972</td>\n",
       "      <td>2.997236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>networks</td>\n",
       "      <td>3.449347</td>\n",
       "      <td>9.767373</td>\n",
       "      <td>2.831658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>learning</td>\n",
       "      <td>9.571162</td>\n",
       "      <td>17.632242</td>\n",
       "      <td>1.842226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>deep</td>\n",
       "      <td>0.139838</td>\n",
       "      <td>8.946511</td>\n",
       "      <td>63.977491</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   word  old freq (%)  new freq (%)       ratio\n",
       "0         deep learning      0.000000      2.169210         inf\n",
       "1             detection      2.377253      4.637724    1.950875\n",
       "2  convolutional neural      0.000000      2.290710         inf\n",
       "3                 image      2.843381      5.746036    2.020846\n",
       "4       neural networks      1.258546      4.868869    3.868647\n",
       "5         convolutional      0.015538      3.834642  246.797570\n",
       "6                neural      3.107520      9.313972    2.997236\n",
       "7              networks      3.449347      9.767373    2.831658\n",
       "8              learning      9.571162     17.632242    1.842226\n",
       "9                  deep      0.139838      8.946511   63.977491"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(\n",
    "    [(keys[i], 100*old_counts[i], 100*new_counts[i], new_counts[i]/old_counts[i]) for i in arg_diff[-10:]],\n",
    "    columns = ['word', 'old freq (%)', 'new freq (%)', 'ratio']\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Words such as:\n",
    "* logic\n",
    "* programming\n",
    "* algorithms\n",
    "* evolutionary\n",
    "\n",
    "used to appear in between 1% and 5% or titles.\n",
    "Their usage has generally been cut by a half or a third.\n",
    "\n",
    "Words (and word pairs) such as:\n",
    "* deep learning\n",
    "* deep\n",
    "* convolutional\n",
    "\n",
    "have increased dramatically. In some cases they appear over 100 times more often than after 2012 compared to before 2012."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Examining popular media\n",
    "\n",
    "Let's compare the trends from arxiv to how the word AI is trending in guardian news articles. \n",
    "The trend is remarkably similar to the deep learning trend in accademic publications."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## Uncomment this cell to download the guardian data from scratch, you will need an API key\n",
    "# ! python guardian.py > guardian.jsonlines\n",
    "# guardian_data = []\n",
    "# guardian_data_df = pd.read_json('guardian.jsonlines', lines=True, orient='records')\n",
    "# guardian_data_df = guardian_data_df[guardian_data_df['year'] < 2018]\n",
    "# guardian_data_df['year'] = guardian_data_df['webPublicationDate'].str.split('-').apply(lambda x: int(x[0]))\n",
    "# for year, group in guardian_data_df.groupby('year'):\n",
    "#     guardian_data.append({'year': year, 'ai': sum(group['webTitle'].str.contains('AI') > 0), 'total': len(group)})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAElCAYAAAAfhqICAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nu5lZBf0vI2U/y1l21DwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a4f7160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "guardian_data = pd.DataFrame([  \n",
    "   {'ai': 1, 'year': 2010, 'total': 7201},\n",
    "   {'ai':2, 'year': 2011, 'total': 8117},\n",
    "   {'ai':1, 'year': 2012, 'total': 8376},\n",
    "   {'ai':3, 'year': 2013, 'total': 9874},\n",
    "   {'ai':7, 'year': 2014, 'total': 10569},\n",
    "   {'ai':16, 'year': 2015, 'total': 9535},\n",
    "   {'ai':51, 'year': 2016, 'total': 9366},\n",
    "   {'ai':60, 'year': 2017, 'total': 7272},\n",
    "])\n",
    "\n",
    "plt.bar(guardian_data['year'], 100 * guardian_data['ai'] / guardian_data['total'])\n",
    "plt.xlabel('year')\n",
    "plt.title('Percentage of guardian technology \\n headlines containing AI')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ten Guardian Headlines Chosen at Random\n",
    "\n",
    "[\"2017-10-18T17:00:02Z\",\"'It's able to create knowledge itself': Google unveils AI that learns on its own\"\n",
    "\n",
    "[\"2016-02-13T15:00:07Z\",\"Would you bet against sex robots? AI 'could leave half of world unemployed'\"\n",
    "\n",
    "[\"2017-01-01T18:11:43Z\",\"Algorithms: AI’s creepy control must be open to inspection\"\n",
    "\n",
    "[\"2016-12-25T12:00:13Z\",\"AI, self-driving cars and cyberwar – the tech trends to watch for in 2017\"\n",
    "\n",
    "[\"2017-05-29T06:00:18Z\",\"Sexist robots can be stopped by women who work in AI\"\n",
    "\n",
    "[\"2017-11-29T06:00:27Z\",\"The Robot Will See You Now review – it appears even therapists could lose their jobs to AI\"\n",
    "\n",
    "[\"2016-10-12T08:21:09Z\",\"Video games where people matter? The strange future of emotional AI\"\n",
    "\n",
    "[\"2017-10-31T15:00:08Z\",\"Horror fiction by numbers: my not-so-shocking AI collaboration\"\n",
    "\n",
    "[\"2011-11-02T09:00:00Z\",\"AI scientists want to make gods. Should that worry us? | Wendy M Grossman\"\n",
    "\n",
    "[\"2016-08-30T13:00:00Z\",\"The rise of robots: forget evil AI – the real risk is far more insidious\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusions\n",
    "\n",
    "The computer science AI landscape has transformed since 2012.\n",
    "Pre 2012 there were a braod range of topics such as `fuzzy logic`, `logic programming`, `evolutionary algorithms`.\n",
    "There each had a presence in 1%-5% of publications.\n",
    "Frequently domain specific algorithms were published that solve niche problems such as face detection or optimal searches.\n",
    "\n",
    "Post 2012 the landscape has become much more monolithic with one topic, deep learning, dominating the landscape.\n",
    "Deep learning has come from obscurity to ubiquity, the subject of approximately one in five publications whilst still growings in its share.\n",
    "Following the incredibly rapid trends demonstated above deep learning is on course to saturate AI research in the next few years, although this is far from a certainty.\n",
    "\n",
    "The data from the guardian strongly suggests that the increase in press overage of AI is being by breakthroughs in deep learning.\n",
    "\n",
    "When reading about AI revolutions in the press we need to remember in reality it is a deep learning revolution.\n",
    "Deep learning is not a silver bullet, while it will continue to make huge strides in some areas (very high dimensional proplems with huge datasets) it is unliekly to have any impact in others.\n",
    "\n",
    "Deep leaning is applied to large datasets for high dimesnional problems such as image recognition, text to speach and translation. This is reflected in advances in products like Google Translate, Siri, Alexa and visual search tools. Deep leaning is unlikely to have anywhere near as large impact for low dimensional problems or problems with modest datasets where traditional buesness analytics and stastistics are used.\n",
    "\n",
    "If we do not unsterstand the strengths and weeknesses of deep learning and communicate them to manage expectations we risk a damaging gap between expectation and reality, between researchers and journalists which could be harmful for future progress. "
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}