Small_Scale_Transcripts_And_Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ae96b5f4-73de-4b82-8bfc-80ff3c7024d4",
   "metadata": {},
   "source": [
    "# Small_Scale_Transcripts_And_Analysis.ipynb\n",
    "This notebook:\n",
    "* Maps the podcast ids to their annotation labels.\n",
    "* Obtains the diff metric across the annotators for the whole set, the non-scripted set, and the scripted set of podcasts.\n",
    "* Writes in the WhisperX, Google ASR, GroundTruth-Min, GroundTruth-Max, and GroundTruth-netural podcasts into a df. \n",
    "* Saves as csv."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "1567656b-26ea-4d25-873c-183ee0a97abe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nums</th>\n",
       "      <th>ids</th>\n",
       "      <th>labels</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1ll5WGWjWANfGKHZfrpL5A</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1wHRMsWVurmo56xd1AcSfe</td>\n",
       "      <td>scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3i4qZ6FqzXdj8UiMeuaX3q</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4YOq90opfiBsc3MwflzFbJ</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>5a0dxl6w7wh5SGAbNpxhu4</td>\n",
       "      <td>scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>5TpkTe5mlb5xx1UT8OvZ6l</td>\n",
       "      <td>scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>6jyaOoQ6aF3TB8z93Honak</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>6MaijKHa5MU0MsdS0yVZHE</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>6URm69QBcvJpqoGiTE1nyy</td>\n",
       "      <td>not-scripted</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>7n6zBPeHcH1t3MUvBNIvDa</td>\n",
       "      <td>scripted</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   nums                     ids        labels\n",
       "1     1  1ll5WGWjWANfGKHZfrpL5A  not-scripted\n",
       "2     2  1wHRMsWVurmo56xd1AcSfe      scripted\n",
       "3     3  3i4qZ6FqzXdj8UiMeuaX3q  not-scripted\n",
       "4     4  4YOq90opfiBsc3MwflzFbJ  not-scripted\n",
       "5     5  5a0dxl6w7wh5SGAbNpxhu4      scripted\n",
       "6     6  5TpkTe5mlb5xx1UT8OvZ6l      scripted\n",
       "7     7  6jyaOoQ6aF3TB8z93Honak  not-scripted\n",
       "8     8  6MaijKHa5MU0MsdS0yVZHE  not-scripted\n",
       "9     9  6URm69QBcvJpqoGiTE1nyy  not-scripted\n",
       "10   10  7n6zBPeHcH1t3MUvBNIvDa      scripted"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import utils_general\n",
    "\n",
    "import os\n",
    "import difflib\n",
    "import pandas as pd\n",
    "import json\n",
    "import string\n",
    "import re\n",
    "\n",
    "from urlextract import URLExtract\n",
    "\n",
    "# useful global variables\n",
    "nums_to_ids = {\n",
    "    \"1\" : \"1ll5WGWjWANfGKHZfrpL5A\",\n",
    "    \"2\" : \"1wHRMsWVurmo56xd1AcSfe\",\n",
    "    \"3\" : \"3i4qZ6FqzXdj8UiMeuaX3q\",\n",
    "    \"4\" : \"4YOq90opfiBsc3MwflzFbJ\",\n",
    "    \"5\" : \"5a0dxl6w7wh5SGAbNpxhu4\",\n",
    "    \"6\" : \"5TpkTe5mlb5xx1UT8OvZ6l\",\n",
    "    \"7\" : \"6jyaOoQ6aF3TB8z93Honak\",\n",
    "    \"8\" : \"6MaijKHa5MU0MsdS0yVZHE\",\n",
    "    \"9\" : \"6URm69QBcvJpqoGiTE1nyy\",\n",
    "    \"10\" : \"7n6zBPeHcH1t3MUvBNIvDa\"\n",
    "}\n",
    "\n",
    "nums_to_labels = {\n",
    "    \"1\" : \"not-scripted\",\n",
    "    \"2\" : \"scripted\",\n",
    "    \"3\" : \"not-scripted\",\n",
    "    \"4\" : \"not-scripted\",\n",
    "    \"5\" : \"scripted\",\n",
    "    \"6\" : \"scripted\",\n",
    "    \"7\" : \"not-scripted\",\n",
    "    \"8\" : \"not-scripted\",\n",
    "    \"9\" : \"not-scripted\",\n",
    "    \"10\" : \"scripted\"\n",
    "}\n",
    "\n",
    "annotators = [\"Anwesha\", \"Eric\", \"GroundTruth\", \"Soohwan\"]\n",
    "\n",
    "# initialize df\n",
    "df = pd.DataFrame(index=range(1,10+1))\n",
    "\n",
    "df[\"nums\"] = range(1, 10 + 1)\n",
    "df[\"nums\"] = df[\"nums\"].astype(str)\n",
    "\n",
    "df[\"ids\"] = df[\"nums\"].map(nums_to_ids)\n",
    "df[\"labels\"] = df[\"nums\"].map(nums_to_labels)\n",
    "\n",
    "display(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ecd4f5ce-b9af-4a25-971a-3f1f93df4574",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nums</th>\n",
       "      <th>ids</th>\n",
       "      <th>labels</th>\n",
       "      <th>total_diff_avg</th>\n",
       "      <th>left1_2</th>\n",
       "      <th>right1_2</th>\n",
       "      <th>total1_2</th>\n",
       "      <th>left2_3</th>\n",
       "      <th>right2_3</th>\n",
       "      <th>total2_3</th>\n",
       "      <th>left1_3</th>\n",
       "      <th>right1_3</th>\n",
       "      <th>total1_3</th>\n",
       "      <th>len1</th>\n",
       "      <th>len2</th>\n",
       "      <th>len3</th>\n",
       "      <th>len_avg</th>\n",
       "      <th>percent_diff</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1ll5WGWjWANfGKHZfrpL5A</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>38.00</td>\n",
       "      <td>27.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>56.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>267.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>267.0</td>\n",
       "      <td>267.67</td>\n",
       "      <td>14.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1wHRMsWVurmo56xd1AcSfe</td>\n",
       "      <td>scripted</td>\n",
       "      <td>3.33</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>270.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>269.33</td>\n",
       "      <td>1.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3i4qZ6FqzXdj8UiMeuaX3q</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>51.33</td>\n",
       "      <td>27.0</td>\n",
       "      <td>40.0</td>\n",
       "      <td>67.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>18.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>363.0</td>\n",
       "      <td>376.0</td>\n",
       "      <td>377.0</td>\n",
       "      <td>372.00</td>\n",
       "      <td>13.80</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4YOq90opfiBsc3MwflzFbJ</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>26.67</td>\n",
       "      <td>19.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>36.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>295.0</td>\n",
       "      <td>293.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>294.67</td>\n",
       "      <td>9.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>5a0dxl6w7wh5SGAbNpxhu4</td>\n",
       "      <td>scripted</td>\n",
       "      <td>24.00</td>\n",
       "      <td>16.0</td>\n",
       "      <td>19.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>26.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>270.0</td>\n",
       "      <td>273.0</td>\n",
       "      <td>269.0</td>\n",
       "      <td>270.67</td>\n",
       "      <td>8.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>5TpkTe5mlb5xx1UT8OvZ6l</td>\n",
       "      <td>scripted</td>\n",
       "      <td>16.67</td>\n",
       "      <td>11.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>306.0</td>\n",
       "      <td>308.0</td>\n",
       "      <td>306.0</td>\n",
       "      <td>306.67</td>\n",
       "      <td>5.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>6jyaOoQ6aF3TB8z93Honak</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>39.33</td>\n",
       "      <td>33.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>58.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>37.0</td>\n",
       "      <td>321.0</td>\n",
       "      <td>313.0</td>\n",
       "      <td>318.0</td>\n",
       "      <td>317.33</td>\n",
       "      <td>12.39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>6MaijKHa5MU0MsdS0yVZHE</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>26.67</td>\n",
       "      <td>19.0</td>\n",
       "      <td>21.0</td>\n",
       "      <td>40.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>16.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>256.0</td>\n",
       "      <td>258.0</td>\n",
       "      <td>258.0</td>\n",
       "      <td>257.33</td>\n",
       "      <td>10.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>6URm69QBcvJpqoGiTE1nyy</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>76.67</td>\n",
       "      <td>35.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>106.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>48.0</td>\n",
       "      <td>31.0</td>\n",
       "      <td>45.0</td>\n",
       "      <td>76.0</td>\n",
       "      <td>282.0</td>\n",
       "      <td>318.0</td>\n",
       "      <td>296.0</td>\n",
       "      <td>298.67</td>\n",
       "      <td>25.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>7n6zBPeHcH1t3MUvBNIvDa</td>\n",
       "      <td>scripted</td>\n",
       "      <td>15.33</td>\n",
       "      <td>11.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>11.0</td>\n",
       "      <td>288.0</td>\n",
       "      <td>289.0</td>\n",
       "      <td>289.0</td>\n",
       "      <td>288.67</td>\n",
       "      <td>5.31</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   nums                     ids        labels  total_diff_avg  left1_2  \\\n",
       "1     1  1ll5WGWjWANfGKHZfrpL5A  not-scripted           38.00     27.0   \n",
       "2     2  1wHRMsWVurmo56xd1AcSfe      scripted            3.33      2.0   \n",
       "3     3  3i4qZ6FqzXdj8UiMeuaX3q  not-scripted           51.33     27.0   \n",
       "4     4  4YOq90opfiBsc3MwflzFbJ  not-scripted           26.67     19.0   \n",
       "5     5  5a0dxl6w7wh5SGAbNpxhu4      scripted           24.00     16.0   \n",
       "6     6  5TpkTe5mlb5xx1UT8OvZ6l      scripted           16.67     11.0   \n",
       "7     7  6jyaOoQ6aF3TB8z93Honak  not-scripted           39.33     33.0   \n",
       "8     8  6MaijKHa5MU0MsdS0yVZHE  not-scripted           26.67     19.0   \n",
       "9     9  6URm69QBcvJpqoGiTE1nyy  not-scripted           76.67     35.0   \n",
       "10   10  7n6zBPeHcH1t3MUvBNIvDa      scripted           15.33     11.0   \n",
       "\n",
       "    right1_2  total1_2  left2_3  right2_3  total2_3  left1_3  right1_3  \\\n",
       "1       29.0      56.0      4.0       2.0       6.0     26.0      26.0   \n",
       "2        3.0       5.0      2.0       1.0       3.0      1.0       1.0   \n",
       "3       40.0      67.0     17.0      18.0      35.0     19.0      33.0   \n",
       "4       17.0      36.0      7.0      10.0      17.0     13.0      14.0   \n",
       "5       19.0      35.0     15.0      11.0      26.0      6.0       5.0   \n",
       "6       13.0      24.0      8.0       6.0      14.0      6.0       6.0   \n",
       "7       25.0      58.0      9.0      14.0      23.0     20.0      17.0   \n",
       "8       21.0      40.0      8.0       8.0      16.0     11.0      13.0   \n",
       "9       71.0     106.0     35.0      13.0      48.0     31.0      45.0   \n",
       "10      12.0      23.0      6.0       6.0      12.0      5.0       6.0   \n",
       "\n",
       "    total1_3   len1   len2   len3  len_avg  percent_diff  \n",
       "1       52.0  267.0  269.0  267.0   267.67         14.20  \n",
       "2        2.0  269.0  270.0  269.0   269.33          1.24  \n",
       "3       52.0  363.0  376.0  377.0   372.00         13.80  \n",
       "4       27.0  295.0  293.0  296.0   294.67          9.05  \n",
       "5       11.0  270.0  273.0  269.0   270.67          8.87  \n",
       "6       12.0  306.0  308.0  306.0   306.67          5.44  \n",
       "7       37.0  321.0  313.0  318.0   317.33         12.39  \n",
       "8       24.0  256.0  258.0  258.0   257.33         10.36  \n",
       "9       76.0  282.0  318.0  296.0   298.67         25.67  \n",
       "10      11.0  288.0  289.0  289.0   288.67          5.31  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>labels</th>\n",
       "      <th colspan=\"6\" halign=\"left\">percent_diff</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "      <th>std</th>\n",
       "      <th>median</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>not-scripted</td>\n",
       "      <td>14.245</td>\n",
       "      <td>9.05</td>\n",
       "      <td>25.67</td>\n",
       "      <td>6</td>\n",
       "      <td>5.937221</td>\n",
       "      <td>13.095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>scripted</td>\n",
       "      <td>5.215</td>\n",
       "      <td>1.24</td>\n",
       "      <td>8.87</td>\n",
       "      <td>4</td>\n",
       "      <td>3.120860</td>\n",
       "      <td>5.375</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         labels percent_diff                                     \n",
       "                        mean   min    max count       std  median\n",
       "0  not-scripted       14.245  9.05  25.67     6  5.937221  13.095\n",
       "1      scripted        5.215  1.24   8.87     4  3.120860   5.375"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def get_diff_counts(compare_diff):\n",
    "    plus_count = 0\n",
    "    minus_count = 0\n",
    "    total_count = 0\n",
    "\n",
    "    for i in compare_diff:\n",
    "        word_diff = i.split()\n",
    "        \n",
    "        if len(word_diff) == 2:\n",
    "\n",
    "            if word_diff[0] == \"+\":\n",
    "                plus_count += 1\n",
    "\n",
    "            if word_diff[0] == \"-\":\n",
    "                minus_count += 1\n",
    "                \n",
    "            # we ignore \"?\", this is why: https://stackoverflow.com/questions/28452835/python-difflib-reporting-unwanted-difference\n",
    "\n",
    "    total_count = plus_count + minus_count\n",
    "\n",
    "    return minus_count, plus_count, total_count\n",
    "\n",
    "def find_string_differences(str1, str2, str3):\n",
    "    d = difflib.Differ()\n",
    "    diff1_2 = list(d.compare(str1, str2))\n",
    "    diff2_3 = list(d.compare(str2, str3))\n",
    "    diff1_3 = list(d.compare(str1, str3))\n",
    "\n",
    "    return diff1_2, diff2_3, diff1_3\n",
    "\n",
    "# for the 10 files\n",
    "for i in range(1,11):\n",
    "    \n",
    "    string1 = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", annotators[0], str(i)+\".txt\")).replace(\"’\",\"'\").split()\n",
    "    string2 = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", annotators[1], str(i)+\".txt\")).split()\n",
    "    string3 = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", annotators[2], str(i)+\".txt\")).split()\n",
    "    \n",
    "    diff1_2, diff2_3, diff1_3 = find_string_differences(string1, string2, string3)\n",
    "    \n",
    "    left1_2, right1_2, total1_2 = get_diff_counts(diff1_2)\n",
    "    left2_3, right2_3, total2_3 = get_diff_counts(diff2_3)\n",
    "    left1_3, right1_3, total1_3 = get_diff_counts(diff1_3)\n",
    "    \n",
    "    df.loc[i, \"total_diff_avg\"] = round(float(total1_2 + total2_3 + total1_3)/ 3.0, 2)\n",
    "    \n",
    "    df.loc[i, \"left1_2\"] = left1_2\n",
    "    df.loc[i, \"right1_2\"] = right1_2\n",
    "    df.loc[i, \"total1_2\"] = total1_2\n",
    "    \n",
    "    df.loc[i, \"left2_3\"] = left2_3\n",
    "    df.loc[i, \"right2_3\"] = right2_3\n",
    "    df.loc[i, \"total2_3\"] = total2_3\n",
    "    \n",
    "    df.loc[i, \"left1_3\"] = left1_3\n",
    "    df.loc[i, \"right1_3\"] = right1_3\n",
    "    df.loc[i, \"total1_3\"] = total1_3\n",
    "    \n",
    "    df.loc[i, \"len1\"] = len(string1)\n",
    "    df.loc[i, \"len2\"] = len(string2)\n",
    "    df.loc[i, \"len3\"] = len(string3)\n",
    "    df.loc[i, \"len_avg\"] = round(float(len(string1) + len(string2) + len(string3))/ 3.0, 2)\n",
    "    \n",
    "df[\"percent_diff\"] = round((df[\"total_diff_avg\"]/df[\"len_avg\"])*100,2)\n",
    "\n",
    "# display percent diff results\n",
    "display(df)\n",
    "display(df.groupby(\"labels\").agg({\"percent_diff\": [\"mean\", \"min\", \"max\", \"count\", \"std\", \"median\"]}).reset_index())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "2141bf2d-f258-4534-a360-b700b31620e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Google ASR functions\n",
    "def clean_urls(text):\n",
    "    extractor = URLExtract()\n",
    "    urls = extractor.find_urls(text)\n",
    "    for url in urls:\n",
    "        text = text.replace(url, \"\")\n",
    "    return text\n",
    "\n",
    "# modified from https://github.com/potsawee/podcast_trec2020/blob/main/data/processor.py\n",
    "# also performs some basic cleaning\n",
    "def get_transcript_text_from_json_asr_file(json_asr_file):\n",
    "    transcript_list = []\n",
    "    with open(json_asr_file) as f:\n",
    "        transcript_dict = json.loads(f.read())\n",
    "        \n",
    "        results_list = [r for r in transcript_dict[\"results\"]]\n",
    "        last_result = results_list[-1]\n",
    "        \n",
    "        for word_dict in last_result[\"alternatives\"][0][\"words\"]:\n",
    "            endTime = float(word_dict[\"endTime\"].replace(\"s\",\"\"))\n",
    "            if endTime <= (60 * 2):  # truncates to 2 minutes\n",
    "                transcript_list.append(word_dict[\"word\"])\n",
    "        \n",
    "        transcript_string = \" \".join(transcript_list)\n",
    "        \n",
    "        # clean the transcripts the same way as the descriptions\n",
    "        transcript_string = clean_urls(transcript_string)\n",
    "        transcript_string = transcript_string.encode(\"ascii\", \"ignore\").decode()\n",
    "        \n",
    "        if transcript_string != \"\" and transcript_string[-1] not in string.punctuation:\n",
    "            transcript_string += \".\"\n",
    "        \n",
    "        return transcript_string"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "a118a3fd-44aa-4351-ba40-e97b54210924",
   "metadata": {},
   "outputs": [],
   "source": [
    "# WhisperX functions\n",
    "def get_ascii_text(dictionary):\n",
    "    text = \"\"\n",
    "    \n",
    "    # read text from json object\n",
    "    for t in dictionary[\"segments\"]:\n",
    "        text += t[\"text\"]\n",
    "    text = text.strip()\n",
    "    \n",
    "    # only allow ascii characters\n",
    "    text = text.encode(\"ascii\", \"ignore\").decode()\n",
    "        \n",
    "    return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "24ab0206-f41e-4149-bf3e-4d2635e69dbc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# filter down df\n",
    "df = df[[\"nums\",\"ids\",\"labels\",\"percent_diff\"]]\n",
    "\n",
    "# merge with metadata_df\n",
    "metadata_df = pd.read_csv(utils_general.PATH_TO_2020_TEST_DF, sep=\"\\t\")\n",
    "df = pd.merge(df, metadata_df, left_on=\"ids\", right_on=\"episode_filename_prefix\", how=\"left\")\n",
    "df.index += 1\n",
    "\n",
    "# for the 10 files, read in the GroundTruth, GroundTruth-punctuated, GoogleASR, and WhisperX transcripts\n",
    "for index, row in df.iterrows():\n",
    "    \n",
    "    # read in the GroundTruth transcriptions\n",
    "    df.loc[index, \"GroundTruth\"] = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", \"GroundTruth\", str(index)+\".txt\"))\n",
    "    \n",
    "    # read in the GroundTruth punctuated transcriptions\n",
    "    df.loc[index, \"GroundTruth-max\"] = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", \"GroundTruth-max\", str(index)+\".txt\"))\n",
    "    df.loc[index, \"GroundTruth-neutral\"] = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", \"GroundTruth-neutral\", str(index)+\".txt\"))\n",
    "    df.loc[index, \"GroundTruth-min\"] = utils_general.read_file(os.path.join(utils_general.PATH_TO_PROJECT, \"annotations\", \"GroundTruth-min\", str(index)+\".txt\"))\n",
    "    \n",
    "    # read in the GoogleASR transcriptions\n",
    "    show_filename_prefix = row[\"show_filename_prefix\"]\n",
    "    episode_filename_prefix = row[\"episode_filename_prefix\"]\n",
    "    dir1 = show_filename_prefix.split(\"_\")[1][0].upper()\n",
    "    dir2 = show_filename_prefix.split(\"_\")[1][1].upper()\n",
    "    json_asr_file = os.path.join(\"/data2/maria/Spotify-Podcasts/podcasts-no-audio-13GB/TREC/spotify-podcasts-2020/podcasts-transcripts-summarization-testset\", dir1, dir2, show_filename_prefix, episode_filename_prefix+\".json\")\n",
    "    df.loc[index, \"GoogleASR\"] = get_transcript_text_from_json_asr_file(json_asr_file)\n",
    "    \n",
    "    # read in the WhisperX transcriptions\n",
    "    episode_id = row[\"ids\"]\n",
    "    with open(f\"/data1/maria/Spotify-Podcasts/test-2min-whisperx-dir/{episode_id}/transcript.json\") as f:\n",
    "        dictionary = json.loads(f.read())\n",
    "    df.loc[index, \"WhisperX\"] = get_ascii_text(dictionary)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "6ec4f736-2422-4039-9583-491ff2f29654",
   "metadata": {},
   "outputs": [],
   "source": [
    "def remove_punctuation_and_lower(text):\n",
    "    text = text.lower()\n",
    "    for p in \"!\\\"#$%&()*+,-./:;<=>?@[\\]^_`{|}~\":  # removed ' character from string.punctuation to keep ' in words like \"you're\"\n",
    "        text = text.replace(p, \"\")\n",
    "    \n",
    "    return text\n",
    "\n",
    "df[\"rpl_GroundTruth\"] = df[\"GroundTruth\"].apply(remove_punctuation_and_lower) \n",
    "df[\"rpl_GoogleASR\"] = df[\"GoogleASR\"].apply(remove_punctuation_and_lower)\n",
    "df[\"rpl_WhisperX\"] = df[\"WhisperX\"].apply(remove_punctuation_and_lower)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "b26707f5-72b7-437a-8b5d-76e438f5131a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nums</th>\n",
       "      <th>ids</th>\n",
       "      <th>labels</th>\n",
       "      <th>percent_diff</th>\n",
       "      <th>show_uri</th>\n",
       "      <th>show_name</th>\n",
       "      <th>show_description</th>\n",
       "      <th>publisher</th>\n",
       "      <th>language</th>\n",
       "      <th>rss_link</th>\n",
       "      <th>...</th>\n",
       "      <th>episode_filename_prefix</th>\n",
       "      <th>GroundTruth</th>\n",
       "      <th>GroundTruth-max</th>\n",
       "      <th>GroundTruth-neutral</th>\n",
       "      <th>GroundTruth-min</th>\n",
       "      <th>GoogleASR</th>\n",
       "      <th>WhisperX</th>\n",
       "      <th>rpl_GroundTruth</th>\n",
       "      <th>rpl_GoogleASR</th>\n",
       "      <th>rpl_WhisperX</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1ll5WGWjWANfGKHZfrpL5A</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>14.20</td>\n",
       "      <td>spotify:show:5HLsz7WFjW8hzJurMDdozi</td>\n",
       "      <td>Granger Smith Podcast</td>\n",
       "      <td>American, Texan, father, husband, musician, Gr...</td>\n",
       "      <td>Granger Smith</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/11170a28/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>1ll5WGWjWANfGKHZfrpL5A</td>\n",
       "      <td>if you're ever in deep deep grief and you wond...</td>\n",
       "      <td>If you're ever in deep, deep grief, and you wo...</td>\n",
       "      <td>If you're ever in deep, deep grief and you won...</td>\n",
       "      <td>If you're ever in deep, deep grief and you won...</td>\n",
       "      <td>If you're ever in deep deep grief and you wond...</td>\n",
       "      <td>If you're ever in deep, deep grief and you won...</td>\n",
       "      <td>if you're ever in deep deep grief and you wond...</td>\n",
       "      <td>if you're ever in deep deep grief and you wond...</td>\n",
       "      <td>if you're ever in deep deep grief and you wond...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1wHRMsWVurmo56xd1AcSfe</td>\n",
       "      <td>scripted</td>\n",
       "      <td>1.24</td>\n",
       "      <td>spotify:show:1g056e2x0Y9AwW6CQF3qA5</td>\n",
       "      <td>Mythology</td>\n",
       "      <td>Myths endure for a reason. This episodic audio...</td>\n",
       "      <td>Parcast Network</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://feeds.megaphone.fm/mythology</td>\n",
       "      <td>...</td>\n",
       "      <td>1wHRMsWVurmo56xd1AcSfe</td>\n",
       "      <td>something to note all myths have many versions...</td>\n",
       "      <td>Something to note: all myths have many version...</td>\n",
       "      <td>Something to note: all myths have many version...</td>\n",
       "      <td>Something to note: all myths have many version...</td>\n",
       "      <td>Something to note all myths have many versions...</td>\n",
       "      <td>Something to note, all myths have many version...</td>\n",
       "      <td>something to note all myths have many versions...</td>\n",
       "      <td>something to note all myths have many versions...</td>\n",
       "      <td>something to note all myths have many versions...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3i4qZ6FqzXdj8UiMeuaX3q</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>13.80</td>\n",
       "      <td>spotify:show:2z1LtWVnflRUONFAo0FADb</td>\n",
       "      <td>Gamecock Central Podcast Network</td>\n",
       "      <td>The Gamecock Central Podcast Network brings yo...</td>\n",
       "      <td>Gamecock Central Podcasts</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://www.spreaker.com/show/957147/episodes/...</td>\n",
       "      <td>...</td>\n",
       "      <td>3i4qZ6FqzXdj8UiMeuaX3q</td>\n",
       "      <td>hello and welcome back into another episode of...</td>\n",
       "      <td>Hello, and welcome back into another episode o...</td>\n",
       "      <td>Hello and welcome back into another episode of...</td>\n",
       "      <td>Hello and welcome back into another episode of...</td>\n",
       "      <td>Hello and welcome back in to another episode o...</td>\n",
       "      <td>Hello and welcome back into another episode of...</td>\n",
       "      <td>hello and welcome back into another episode of...</td>\n",
       "      <td>hello and welcome back in to another episode o...</td>\n",
       "      <td>hello and welcome back into another episode of...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4YOq90opfiBsc3MwflzFbJ</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>9.05</td>\n",
       "      <td>spotify:show:4O6pIMQHq2GkVhwQ0KNdnH</td>\n",
       "      <td>The Frog and The Dragon</td>\n",
       "      <td>A podcast devoted to everything Magic: the Gat...</td>\n",
       "      <td>The Frog &amp; The Dragon</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/1338c918/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>4YOq90opfiBsc3MwflzFbJ</td>\n",
       "      <td>hello everyone and welcome to the frog and the...</td>\n",
       "      <td>Hello, everyone, and welcome to \"The Frog and ...</td>\n",
       "      <td>Hello everyone, and welcome to The Frog and th...</td>\n",
       "      <td>Hello everyone, and welcome to The Frog and Th...</td>\n",
       "      <td>Hello everyone and welcome to frog and the dra...</td>\n",
       "      <td>Hello everyone and welcome to the Frog and the...</td>\n",
       "      <td>hello everyone and welcome to the frog and the...</td>\n",
       "      <td>hello everyone and welcome to frog and the dra...</td>\n",
       "      <td>hello everyone and welcome to the frog and the...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>5a0dxl6w7wh5SGAbNpxhu4</td>\n",
       "      <td>scripted</td>\n",
       "      <td>8.87</td>\n",
       "      <td>spotify:show:21ASCcEXgUlbFSmoqjroZm</td>\n",
       "      <td>Chompers</td>\n",
       "      <td>Make tooth time easy with this morning and nig...</td>\n",
       "      <td>Gimlet</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://feeds.megaphone.fm/chompers</td>\n",
       "      <td>...</td>\n",
       "      <td>5a0dxl6w7wh5SGAbNpxhu4</td>\n",
       "      <td>good morning we hope you're ready to have a su...</td>\n",
       "      <td>Good morning! We hope you're ready to have a s...</td>\n",
       "      <td>Good morning! We hope you're ready to have a s...</td>\n",
       "      <td>Good morning, we hope you're ready to have a s...</td>\n",
       "      <td>Good morning. We hope you're ready to have a s...</td>\n",
       "      <td>Good morning. We hope you're ready to have a s...</td>\n",
       "      <td>good morning we hope you're ready to have a su...</td>\n",
       "      <td>good morning we hope you're ready to have a su...</td>\n",
       "      <td>good morning we hope you're ready to have a su...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>5TpkTe5mlb5xx1UT8OvZ6l</td>\n",
       "      <td>scripted</td>\n",
       "      <td>5.44</td>\n",
       "      <td>spotify:show:4gds4Ip86RDAlAgH4PQuQs</td>\n",
       "      <td>Crime Over Coffee</td>\n",
       "      <td>Listen in each week while true crime enthusias...</td>\n",
       "      <td>Crime Over Coffee</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/b402dc8/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>5TpkTe5mlb5xx1UT8OvZ6l</td>\n",
       "      <td>good morning and welcome to another mini monda...</td>\n",
       "      <td>Good morning, and welcome to another mini Mond...</td>\n",
       "      <td>Good morning, and welcome to another Mini Mond...</td>\n",
       "      <td>Good morning, and welcome to another Mini Mond...</td>\n",
       "      <td>Good morning, and welcome to another mini Mond...</td>\n",
       "      <td>Good morning, and welcome to another mini Mond...</td>\n",
       "      <td>good morning and welcome to another mini monda...</td>\n",
       "      <td>good morning and welcome to another mini monda...</td>\n",
       "      <td>good morning and welcome to another mini monda...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>6jyaOoQ6aF3TB8z93Honak</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>12.39</td>\n",
       "      <td>spotify:show:6JqOLHQRh2G24ItUTlmxu9</td>\n",
       "      <td>Legends Cast - A Podcast About Legends of Rune...</td>\n",
       "      <td>A podcast about the meta, cards, and community...</td>\n",
       "      <td>The Lift</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/c4221b8/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>6jyaOoQ6aF3TB8z93Honak</td>\n",
       "      <td>you're listening to legends cast a podcast abo...</td>\n",
       "      <td>You're listening to Legends Cast, a podcast ab...</td>\n",
       "      <td>You're listening to Legends Cast, a podcast ab...</td>\n",
       "      <td>You're listening to Legends Cast, a podcast ab...</td>\n",
       "      <td>You're listening to Legends cast a podcast abo...</td>\n",
       "      <td>You're listening to Legendscast, a podcast abo...</td>\n",
       "      <td>you're listening to legends cast a podcast abo...</td>\n",
       "      <td>you're listening to legends cast a podcast abo...</td>\n",
       "      <td>you're listening to legendscast a podcast abou...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>6MaijKHa5MU0MsdS0yVZHE</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>10.36</td>\n",
       "      <td>spotify:show:0cTycvoAlE6LSppQHpgr48</td>\n",
       "      <td>Align By Design</td>\n",
       "      <td>Align by Design with Amy Allchurch is the podc...</td>\n",
       "      <td>Amy Allchurch</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/de57c04/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>6MaijKHa5MU0MsdS0yVZHE</td>\n",
       "      <td>welcome to align by design i'm amy allchurch a...</td>\n",
       "      <td>Welcome to Align by Design; I'm Amy Allchurch,...</td>\n",
       "      <td>Welcome to Align by Design, I'm Amy Allchurch,...</td>\n",
       "      <td>Welcome to Align by Design, I'm Amy Allchurch,...</td>\n",
       "      <td>Welcome to 9 by Design. I'm eating all church ...</td>\n",
       "      <td>Welcome to Align by Design. I'm Amy Alchurch, ...</td>\n",
       "      <td>welcome to align by design i'm amy allchurch a...</td>\n",
       "      <td>welcome to 9 by design i'm eating all church a...</td>\n",
       "      <td>welcome to align by design i'm amy alchurch an...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>6URm69QBcvJpqoGiTE1nyy</td>\n",
       "      <td>not-scripted</td>\n",
       "      <td>25.67</td>\n",
       "      <td>spotify:show:40tQwCsD43wEfTZuDuZ66c</td>\n",
       "      <td>Let's All Be Serious</td>\n",
       "      <td>Imagine sitting around a table with three of y...</td>\n",
       "      <td>Banger</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://anchor.fm/s/f7a6444/podcast/rss</td>\n",
       "      <td>...</td>\n",
       "      <td>6URm69QBcvJpqoGiTE1nyy</td>\n",
       "      <td>so is that all they do for icy hot those lying...</td>\n",
       "      <td>So, is that all they do for Icy Hot, those lyi...</td>\n",
       "      <td>So, is that all they do for Icy Hot? Those lyi...</td>\n",
       "      <td>So, is that all they do for Icy Hot, those lyi...</td>\n",
       "      <td>So that all they do for Icy Hot those Lyin mot...</td>\n",
       "      <td>So is that all they do for Icy Hot, those lyin...</td>\n",
       "      <td>so is that all they do for icy hot those lying...</td>\n",
       "      <td>so that all they do for icy hot those lyin mot...</td>\n",
       "      <td>so is that all they do for icy hot those lying...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>7n6zBPeHcH1t3MUvBNIvDa</td>\n",
       "      <td>scripted</td>\n",
       "      <td>5.31</td>\n",
       "      <td>spotify:show:5PFgy6lc8UvYIjDsaaO3dN</td>\n",
       "      <td>Crimes of Passion</td>\n",
       "      <td>Every Wednesday, we tell a love story that tur...</td>\n",
       "      <td>Parcast Network</td>\n",
       "      <td>['en']</td>\n",
       "      <td>https://feeds.megaphone.fm/crimesofpassion</td>\n",
       "      <td>...</td>\n",
       "      <td>7n6zBPeHcH1t3MUvBNIvDa</td>\n",
       "      <td>welcome to podcast crime bites we're offering ...</td>\n",
       "      <td>Welcome to Podcast Crime Bites, where we're of...</td>\n",
       "      <td>Welcome to Podcast Crime Bites. We're offering...</td>\n",
       "      <td>Welcome to Podcast Crime Bites, we're offering...</td>\n",
       "      <td>Welcome to park asked crime bites. We're offer...</td>\n",
       "      <td>Welcome to ParCast Crime Bites. We're offering...</td>\n",
       "      <td>welcome to podcast crime bites we're offering ...</td>\n",
       "      <td>welcome to park asked crime bites we're offeri...</td>\n",
       "      <td>welcome to parcast crime bites we're offering ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   nums                     ids        labels  percent_diff  \\\n",
       "1     1  1ll5WGWjWANfGKHZfrpL5A  not-scripted         14.20   \n",
       "2     2  1wHRMsWVurmo56xd1AcSfe      scripted          1.24   \n",
       "3     3  3i4qZ6FqzXdj8UiMeuaX3q  not-scripted         13.80   \n",
       "4     4  4YOq90opfiBsc3MwflzFbJ  not-scripted          9.05   \n",
       "5     5  5a0dxl6w7wh5SGAbNpxhu4      scripted          8.87   \n",
       "6     6  5TpkTe5mlb5xx1UT8OvZ6l      scripted          5.44   \n",
       "7     7  6jyaOoQ6aF3TB8z93Honak  not-scripted         12.39   \n",
       "8     8  6MaijKHa5MU0MsdS0yVZHE  not-scripted         10.36   \n",
       "9     9  6URm69QBcvJpqoGiTE1nyy  not-scripted         25.67   \n",
       "10   10  7n6zBPeHcH1t3MUvBNIvDa      scripted          5.31   \n",
       "\n",
       "                               show_uri  \\\n",
       "1   spotify:show:5HLsz7WFjW8hzJurMDdozi   \n",
       "2   spotify:show:1g056e2x0Y9AwW6CQF3qA5   \n",
       "3   spotify:show:2z1LtWVnflRUONFAo0FADb   \n",
       "4   spotify:show:4O6pIMQHq2GkVhwQ0KNdnH   \n",
       "5   spotify:show:21ASCcEXgUlbFSmoqjroZm   \n",
       "6   spotify:show:4gds4Ip86RDAlAgH4PQuQs   \n",
       "7   spotify:show:6JqOLHQRh2G24ItUTlmxu9   \n",
       "8   spotify:show:0cTycvoAlE6LSppQHpgr48   \n",
       "9   spotify:show:40tQwCsD43wEfTZuDuZ66c   \n",
       "10  spotify:show:5PFgy6lc8UvYIjDsaaO3dN   \n",
       "\n",
       "                                            show_name  \\\n",
       "1                               Granger Smith Podcast   \n",
       "2                                           Mythology   \n",
       "3                    Gamecock Central Podcast Network   \n",
       "4                             The Frog and The Dragon   \n",
       "5                                            Chompers   \n",
       "6                                   Crime Over Coffee   \n",
       "7   Legends Cast - A Podcast About Legends of Rune...   \n",
       "8                                     Align By Design   \n",
       "9                                Let's All Be Serious   \n",
       "10                                  Crimes of Passion   \n",
       "\n",
       "                                     show_description  \\\n",
       "1   American, Texan, father, husband, musician, Gr...   \n",
       "2   Myths endure for a reason. This episodic audio...   \n",
       "3   The Gamecock Central Podcast Network brings yo...   \n",
       "4   A podcast devoted to everything Magic: the Gat...   \n",
       "5   Make tooth time easy with this morning and nig...   \n",
       "6   Listen in each week while true crime enthusias...   \n",
       "7   A podcast about the meta, cards, and community...   \n",
       "8   Align by Design with Amy Allchurch is the podc...   \n",
       "9   Imagine sitting around a table with three of y...   \n",
       "10  Every Wednesday, we tell a love story that tur...   \n",
       "\n",
       "                    publisher language  \\\n",
       "1               Granger Smith   ['en']   \n",
       "2             Parcast Network   ['en']   \n",
       "3   Gamecock Central Podcasts   ['en']   \n",
       "4       The Frog & The Dragon   ['en']   \n",
       "5                      Gimlet   ['en']   \n",
       "6           Crime Over Coffee   ['en']   \n",
       "7                    The Lift   ['en']   \n",
       "8               Amy Allchurch   ['en']   \n",
       "9                      Banger   ['en']   \n",
       "10            Parcast Network   ['en']   \n",
       "\n",
       "                                             rss_link  ...  \\\n",
       "1            https://anchor.fm/s/11170a28/podcast/rss  ...   \n",
       "2                https://feeds.megaphone.fm/mythology  ...   \n",
       "3   https://www.spreaker.com/show/957147/episodes/...  ...   \n",
       "4            https://anchor.fm/s/1338c918/podcast/rss  ...   \n",
       "5                 https://feeds.megaphone.fm/chompers  ...   \n",
       "6             https://anchor.fm/s/b402dc8/podcast/rss  ...   \n",
       "7             https://anchor.fm/s/c4221b8/podcast/rss  ...   \n",
       "8             https://anchor.fm/s/de57c04/podcast/rss  ...   \n",
       "9             https://anchor.fm/s/f7a6444/podcast/rss  ...   \n",
       "10         https://feeds.megaphone.fm/crimesofpassion  ...   \n",
       "\n",
       "   episode_filename_prefix                                        GroundTruth  \\\n",
       "1   1ll5WGWjWANfGKHZfrpL5A  if you're ever in deep deep grief and you wond...   \n",
       "2   1wHRMsWVurmo56xd1AcSfe  something to note all myths have many versions...   \n",
       "3   3i4qZ6FqzXdj8UiMeuaX3q  hello and welcome back into another episode of...   \n",
       "4   4YOq90opfiBsc3MwflzFbJ  hello everyone and welcome to the frog and the...   \n",
       "5   5a0dxl6w7wh5SGAbNpxhu4  good morning we hope you're ready to have a su...   \n",
       "6   5TpkTe5mlb5xx1UT8OvZ6l  good morning and welcome to another mini monda...   \n",
       "7   6jyaOoQ6aF3TB8z93Honak  you're listening to legends cast a podcast abo...   \n",
       "8   6MaijKHa5MU0MsdS0yVZHE  welcome to align by design i'm amy allchurch a...   \n",
       "9   6URm69QBcvJpqoGiTE1nyy  so is that all they do for icy hot those lying...   \n",
       "10  7n6zBPeHcH1t3MUvBNIvDa  welcome to podcast crime bites we're offering ...   \n",
       "\n",
       "                                      GroundTruth-max  \\\n",
       "1   If you're ever in deep, deep grief, and you wo...   \n",
       "2   Something to note: all myths have many version...   \n",
       "3   Hello, and welcome back into another episode o...   \n",
       "4   Hello, everyone, and welcome to \"The Frog and ...   \n",
       "5   Good morning! We hope you're ready to have a s...   \n",
       "6   Good morning, and welcome to another mini Mond...   \n",
       "7   You're listening to Legends Cast, a podcast ab...   \n",
       "8   Welcome to Align by Design; I'm Amy Allchurch,...   \n",
       "9   So, is that all they do for Icy Hot, those lyi...   \n",
       "10  Welcome to Podcast Crime Bites, where we're of...   \n",
       "\n",
       "                                  GroundTruth-neutral  \\\n",
       "1   If you're ever in deep, deep grief and you won...   \n",
       "2   Something to note: all myths have many version...   \n",
       "3   Hello and welcome back into another episode of...   \n",
       "4   Hello everyone, and welcome to The Frog and th...   \n",
       "5   Good morning! We hope you're ready to have a s...   \n",
       "6   Good morning, and welcome to another Mini Mond...   \n",
       "7   You're listening to Legends Cast, a podcast ab...   \n",
       "8   Welcome to Align by Design, I'm Amy Allchurch,...   \n",
       "9   So, is that all they do for Icy Hot? Those lyi...   \n",
       "10  Welcome to Podcast Crime Bites. We're offering...   \n",
       "\n",
       "                                      GroundTruth-min  \\\n",
       "1   If you're ever in deep, deep grief and you won...   \n",
       "2   Something to note: all myths have many version...   \n",
       "3   Hello and welcome back into another episode of...   \n",
       "4   Hello everyone, and welcome to The Frog and Th...   \n",
       "5   Good morning, we hope you're ready to have a s...   \n",
       "6   Good morning, and welcome to another Mini Mond...   \n",
       "7   You're listening to Legends Cast, a podcast ab...   \n",
       "8   Welcome to Align by Design, I'm Amy Allchurch,...   \n",
       "9   So, is that all they do for Icy Hot, those lyi...   \n",
       "10  Welcome to Podcast Crime Bites, we're offering...   \n",
       "\n",
       "                                            GoogleASR  \\\n",
       "1   If you're ever in deep deep grief and you wond...   \n",
       "2   Something to note all myths have many versions...   \n",
       "3   Hello and welcome back in to another episode o...   \n",
       "4   Hello everyone and welcome to frog and the dra...   \n",
       "5   Good morning. We hope you're ready to have a s...   \n",
       "6   Good morning, and welcome to another mini Mond...   \n",
       "7   You're listening to Legends cast a podcast abo...   \n",
       "8   Welcome to 9 by Design. I'm eating all church ...   \n",
       "9   So that all they do for Icy Hot those Lyin mot...   \n",
       "10  Welcome to park asked crime bites. We're offer...   \n",
       "\n",
       "                                             WhisperX  \\\n",
       "1   If you're ever in deep, deep grief and you won...   \n",
       "2   Something to note, all myths have many version...   \n",
       "3   Hello and welcome back into another episode of...   \n",
       "4   Hello everyone and welcome to the Frog and the...   \n",
       "5   Good morning. We hope you're ready to have a s...   \n",
       "6   Good morning, and welcome to another mini Mond...   \n",
       "7   You're listening to Legendscast, a podcast abo...   \n",
       "8   Welcome to Align by Design. I'm Amy Alchurch, ...   \n",
       "9   So is that all they do for Icy Hot, those lyin...   \n",
       "10  Welcome to ParCast Crime Bites. We're offering...   \n",
       "\n",
       "                                      rpl_GroundTruth  \\\n",
       "1   if you're ever in deep deep grief and you wond...   \n",
       "2   something to note all myths have many versions...   \n",
       "3   hello and welcome back into another episode of...   \n",
       "4   hello everyone and welcome to the frog and the...   \n",
       "5   good morning we hope you're ready to have a su...   \n",
       "6   good morning and welcome to another mini monda...   \n",
       "7   you're listening to legends cast a podcast abo...   \n",
       "8   welcome to align by design i'm amy allchurch a...   \n",
       "9   so is that all they do for icy hot those lying...   \n",
       "10  welcome to podcast crime bites we're offering ...   \n",
       "\n",
       "                                        rpl_GoogleASR  \\\n",
       "1   if you're ever in deep deep grief and you wond...   \n",
       "2   something to note all myths have many versions...   \n",
       "3   hello and welcome back in to another episode o...   \n",
       "4   hello everyone and welcome to frog and the dra...   \n",
       "5   good morning we hope you're ready to have a su...   \n",
       "6   good morning and welcome to another mini monda...   \n",
       "7   you're listening to legends cast a podcast abo...   \n",
       "8   welcome to 9 by design i'm eating all church a...   \n",
       "9   so that all they do for icy hot those lyin mot...   \n",
       "10  welcome to park asked crime bites we're offeri...   \n",
       "\n",
       "                                         rpl_WhisperX  \n",
       "1   if you're ever in deep deep grief and you wond...  \n",
       "2   something to note all myths have many versions...  \n",
       "3   hello and welcome back into another episode of...  \n",
       "4   hello everyone and welcome to the frog and the...  \n",
       "5   good morning we hope you're ready to have a su...  \n",
       "6   good morning and welcome to another mini monda...  \n",
       "7   you're listening to legendscast a podcast abou...  \n",
       "8   welcome to align by design i'm amy alchurch an...  \n",
       "9   so is that all they do for icy hot those lying...  \n",
       "10  welcome to parcast crime bites we're offering ...  \n",
       "\n",
       "[10 rows x 25 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# save the metadata df with the transcripts\n",
    "df.to_csv(\"./csv/small_scale_texts.csv\", header=True)\n",
    "display(df)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:pytorch_env_3.8]",
   "language": "python",
   "name": "conda-env-pytorch_env_3.8-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}