Experimental Results.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Detecting Argumentative Discourse Units with Linguistic Alignment\n",
    "\n",
    "This notebook reproduces the results in our [paper](add_link_here).\n",
    "\n",
    "Note:\n",
    "- The data provided is not text but rather LIWC frequencies for anonymyzed comments, and our annotations\n",
    "- We do not provide anything that would infringe copyright"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pystan\n",
    "\n",
    "from align import logreg, stan, util"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = util.load_json('data')\n",
    "author_baselines = util.load_json('author_baselines')\n",
    "author_means = util.load_json('author_means')\n",
    "illocution_dict = util.load_json('illocution_dict')\n",
    "category_dict = util.load_json('category_dict')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Alignment over Illocutions\n",
    "\n",
    "We first run the stan model to get the posterior distributions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_data = stan.get_model_data(\n",
    "    data, author_baselines, author_means, illocution_dict, \n",
    "    category_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:pystan:COMPILING THE C++ CODE FOR MODEL alignment_illocution_0be51d5572153ad085ca803268f1494d NOW.\n"
     ]
    }
   ],
   "source": [
    "model = pystan.StanModel(file='align/alignment_illocution.stan',\n",
    "                         model_name='alignment_illocution')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_iters = 1000\n",
    "warmup = 500\n",
    "chains = 3\n",
    "thin = 1\n",
    "seed = 42\n",
    "verbose = True\n",
    "pars = ['eta_illocution_align', 'eta_category_align']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "fit = model.sampling(\n",
    "    data=model_data, iter=n_iters, warmup=warmup, chains=chains, \n",
    "    thin=thin, seed=seed, verbose=verbose, pars=pars)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'n_eff': True,\n",
       " 'Rhat': True,\n",
       " 'divergence': True,\n",
       " 'treedepth': True,\n",
       " 'energy': True}"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pystan.check_hmc_diagnostics(fit)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "posteriors = stan.get_posterior_samples(fit, illocution_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plotting the distributions, as per our paper:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 432x1152 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "stan.plot_dists(posteriors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pairwise significance tests for difference of means. We round the p-values to two decimal places. We can see from the matrix displayed that differences are mostly significant with a handful of high p-values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.81, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.51, 0.  , 0.33, 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.52, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.63, 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.18, 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.01, 0.  , 0.03, 0.02, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.52, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.02, 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.26, 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.51, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.1 , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.18, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.18, 0.  , 0.45, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.33, 0.  , 0.  , 0.  , 0.  , 0.1 , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.02, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.01, 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.5 , 0.  ,\n",
       "        0.1 , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.01, 0.  , 0.  , 0.  , 0.18, 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.55, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.81, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.03, 0.  , 0.  , 0.  , 0.45, 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.55, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.02, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  ,\n",
       "        0.07, 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.63, 0.  , 0.26, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.01, 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.05, 0.  , 0.09,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.05, 0.  , 0.  , 0.68,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.5 ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.31, 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.09, 0.68, 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.1 ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.07, 0.  , 0.  , 0.  , 0.31, 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ],\n",
       "       [0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ,\n",
       "        0.  , 0.  , 0.  ]])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stan.pairwise_t(posteriors)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The comparison of pure and assertive questioning, our motivating example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "stan.compare_dists(\n",
    "    posteriors, 'questioning-assertive', 'questioning-pure')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Detecting Argumentative Discourse Acts\n",
    "\n",
    "Logistic regression on each discourse act in each comment.\n",
    "\n",
    "Note:\n",
    "- Here we just supply the already bagged vectors as data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_glove = util.load_npy('x_glove')\n",
    "y = logreg.get_y(data, illocution_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GloVe Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_test = 200\n",
    "x_glove_train = x_glove[0:-n_test, :]\n",
    "y_train = y[0:-n_test]\n",
    "x_glove_test = x_glove[-n_test:, :]\n",
    "y_test = y[-n_test:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 25/25 [11:16<00:00, 28.43s/it]\n"
     ]
    }
   ],
   "source": [
    "glove_results = logreg.evaluate_all(\n",
    "    x_glove_train, y_train, x_glove_test, y_test,\n",
    "    illocution_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### GloVe + Alignment Scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 800/800 [00:00<00:00, 10934.80it/s]\n"
     ]
    }
   ],
   "source": [
    "x_align = logreg.get_x_align(\n",
    "    data, category_dict, author_baselines, author_means)\n",
    "x_glove_align = np.concatenate([x_glove, x_align], axis=1)\n",
    "x_glove_align_train = x_glove_align[0:-n_test, :]\n",
    "x_glove_align_test = x_glove_align[-n_test:, :]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 25/25 [27:34<00:00, 65.06s/it] \n"
     ]
    }
   ],
   "source": [
    "glove_align_results = logreg.evaluate_all(\n",
    "    x_glove_align_train, y_train, x_glove_align_test, y_test,\n",
    "    illocution_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparing Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.684\t0.684\t0.000\targuing\n",
      "0.560\t0.640\t0.081\tasserting\n",
      "0.497\t0.497\t0.000\tquestioning-rhetorical\n",
      "0.497\t0.497\t0.000\tchallenging-pure\n",
      "0.500\t0.500\t0.000\tquestioning-pure\n",
      "0.497\t0.495\t-0.003\tdisagreeing without extra argument\n",
      "0.495\t0.497\t0.003\tquestioning-assertive\n",
      "0.497\t0.490\t-0.008\tchallenging-assertive\n",
      "0.637\t0.650\t0.013\tdisagreeing\n",
      "0.595\t0.628\t0.034\tdisagreeing with extra argument\n",
      "0.495\t0.487\t-0.008\tconceding and disagreeing\n",
      "0.492\t0.535\t0.043\tconceding\n",
      "0.550\t0.630\t0.080\tnon-argumentative\n",
      "0.500\t0.497\t-0.003\tagreeing\n",
      "0.500\t0.500\t0.000\tagreeing without extra argument\n",
      "0.497\t0.497\t0.000\tagreeing with extra argument\n"
     ]
    }
   ],
   "source": [
    "logreg.compare_results(glove_results, glove_align_results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "logreg.plot_comparison(glove_results, glove_align_results)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "stan",
   "language": "python",
   "name": "stan"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}