foodmusic-analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "bb745c66",
   "metadata": {},
   "source": [
    "## Food/Music Analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cc0f153e",
   "metadata": {},
   "source": [
    "This analysis is done without peeking at the learning outcomes."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "327eb0a6",
   "metadata": {},
   "source": [
    "### Setup"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5655b5d6",
   "metadata": {},
   "source": [
    "We import the `coilnlp` package and load in pre-processed dataframes. Note that `numpy`, `pandas`, and `matplotlib.pyplot` have already been imported as `np`, `pd`, and `plt`, respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0ef10201",
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2188b1dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "from coilnlp import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "03dd54a8",
   "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>TOPIC</th>\n",
       "      <th>STATE</th>\n",
       "      <th>DOCNUM</th>\n",
       "      <th>SENTENCE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1430</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>you've probably heard his song with julia mic...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1431</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>some of my favorites of his are     the few t...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1432</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>hey stupid i love you     same room  one ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1433</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>hope you give them a listen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1434</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          TOPIC STATE  DOCNUM  \\\n",
       "1430  foodmusic    us       9   \n",
       "1431  foodmusic    us       9   \n",
       "1432  foodmusic    us       9   \n",
       "1433  foodmusic    us       9   \n",
       "1434  foodmusic    us       9   \n",
       "\n",
       "                                               SENTENCE  \n",
       "1430   you've probably heard his song with julia mic...  \n",
       "1431   some of my favorites of his are     the few t...  \n",
       "1432       hey stupid i love you     same room  one ...  \n",
       "1433                        hope you give them a listen  \n",
       "1434                                                NaN  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sentences\n",
    "covid = pd.read_csv('covid-sent.csv', header = 0)\n",
    "gender = pd.read_csv('gender-sent.csv', header = 0)\n",
    "english = pd.read_csv('english-sent.csv', header = 0)\n",
    "fm = pd.read_csv('foodmusic-sent.csv', header = 0)\n",
    "sent = pd.concat([covid, gender, english, fm], ignore_index = True)\n",
    "sent.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "438fbf3f",
   "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>TOPIC</th>\n",
       "      <th>STATE</th>\n",
       "      <th>DOCNUM</th>\n",
       "      <th>TEXT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>covid</td>\n",
       "      <td>iq</td>\n",
       "      <td>1</td>\n",
       "      <td>the covid 19 pandemic has transformed the cont...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>covid</td>\n",
       "      <td>iq</td>\n",
       "      <td>10</td>\n",
       "      <td>well today me and my group talked about covid ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>covid</td>\n",
       "      <td>iq</td>\n",
       "      <td>11</td>\n",
       "      <td>it was an interesting conversation with my par...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>covid</td>\n",
       "      <td>iq</td>\n",
       "      <td>12</td>\n",
       "      <td>in the educational process covid 19 changes ev...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>covid</td>\n",
       "      <td>iq</td>\n",
       "      <td>13</td>\n",
       "      <td>okay in our group discussion meeting today we ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   TOPIC STATE  DOCNUM                                               TEXT\n",
       "0  covid    iq       1  the covid 19 pandemic has transformed the cont...\n",
       "1  covid    iq      10  well today me and my group talked about covid ...\n",
       "2  covid    iq      11  it was an interesting conversation with my par...\n",
       "3  covid    iq      12  in the educational process covid 19 changes ev...\n",
       "4  covid    iq      13  okay in our group discussion meeting today we ..."
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# documents\n",
    "covid = pd.read_csv('covid-docs.csv', header = 0)\n",
    "gender = pd.read_csv('gender-docs.csv', header = 0)\n",
    "english = pd.read_csv('english-docs.csv', header = 0)\n",
    "fm = pd.read_csv('foodmusic-docs.csv', header = 0)\n",
    "docs = pd.concat([covid, gender, english, fm], ignore_index = True)\n",
    "docs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "42c122ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'1': 51,\n",
       " 'music': 50,\n",
       " 'songs': 32,\n",
       " '2': 28,\n",
       " 'add': 27,\n",
       " 'one': 27,\n",
       " 'like': 27,\n",
       " 'favorite': 24,\n",
       " 'love': 24,\n",
       " 'iraqi': 22,\n",
       " 'bread': 21,\n",
       " 'make': 21,\n",
       " 'really': 20,\n",
       " 'also': 20,\n",
       " '4': 20,\n",
       " 'cup': 19,\n",
       " 'oil': 19,\n",
       " 'put': 19,\n",
       " 'dish': 18,\n",
       " 'song': 18,\n",
       " 'al': 18,\n",
       " 'recipe': 17,\n",
       " 'enjoy': 16,\n",
       " 'ingredients': 16,\n",
       " 'meat': 16,\n",
       " 'listen': 16,\n",
       " 'salt': 15,\n",
       " 'falafel': 14,\n",
       " 'food': 14,\n",
       " 'delicious': 14,\n",
       " 'iraq': 13,\n",
       " \"i'm\": 13,\n",
       " 'time': 13,\n",
       " 'different': 12,\n",
       " 'chicken': 12,\n",
       " 'sauce': 12,\n",
       " 'many': 12,\n",
       " 'pepper': 12,\n",
       " 'minutes': 12,\n",
       " '3': 12,\n",
       " 'culture': 12,\n",
       " 'middle': 11,\n",
       " 'dough': 11,\n",
       " 'oven': 11,\n",
       " 'tsp': 11,\n",
       " 'onions': 11,\n",
       " '5': 10,\n",
       " 'think': 10,\n",
       " 'hope': 10,\n",
       " 'water': 10}"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "word_counts(docs[docs['TOPIC'] == 'foodmusic'], max_words = 50)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "473323a6",
   "metadata": {},
   "source": [
    "The word cloud shows some of the most frequent words in the corpus. We noted other interesting words in the print output: american, (iraq, iraqi), love, like, favorite, bread, song, music, cultur, falafel, chicken, pepper, salt, oil, sauce."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "514bdb24",
   "metadata": {},
   "source": [
    "#### Juxtaposed word clouds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "63f0ce1c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "United States (Food/Music)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=400x200 at 0x16301E31E50>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('United States (Food/Music)')\n",
    "make_word_cloud(docs[(docs['STATE'] == 'us') & (docs['TOPIC'] == 'foodmusic')]).to_image()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "42368cda",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iraq (Food/Music)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=400x200 at 0x1637B506040>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Iraq (Food/Music)')\n",
    "make_word_cloud(docs[(docs['STATE'] == 'iq') & (docs['TOPIC'] == 'foodmusic')]).to_image()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c7d2edc",
   "metadata": {},
   "source": [
    "For the food/music prompt, chocolate, broth, and tradition are used by US students. salt, onion, oil, bread, meat, chicken are used by IQ students. recipe is used by both groups. IQ students now use iraq and iraqi more whereas US students use american. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5679a120",
   "metadata": {},
   "source": [
    "#### Term frequency inverse document frequency"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03dcffa2",
   "metadata": {},
   "source": [
    "Each document focuses on some words and ideas more so than other words and ideas. We can look at the most frequent words for each document."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "995ec1a4",
   "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>Topic</th>\n",
       "      <th>State</th>\n",
       "      <th>DocID</th>\n",
       "      <th>Tfidf 1</th>\n",
       "      <th>Tfidf 2</th>\n",
       "      <th>Tfidf 3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>1</td>\n",
       "      <td>dolma</td>\n",
       "      <td>mom</td>\n",
       "      <td>tbsp</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>10</td>\n",
       "      <td>kebab</td>\n",
       "      <td>grilled</td>\n",
       "      <td>grill</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>11</td>\n",
       "      <td>kibbeh</td>\n",
       "      <td>add</td>\n",
       "      <td>meat</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>12</td>\n",
       "      <td>shawarma</td>\n",
       "      <td>tsp</td>\n",
       "      <td>chicken</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>13</td>\n",
       "      <td>spicy</td>\n",
       "      <td>totally</td>\n",
       "      <td>pumpkin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>14</td>\n",
       "      <td>potatoes</td>\n",
       "      <td>oven</td>\n",
       "      <td>teaspoon</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>15</td>\n",
       "      <td>croissant</td>\n",
       "      <td>cocoa</td>\n",
       "      <td>contains</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>16</td>\n",
       "      <td>cup</td>\n",
       "      <td>sugar</td>\n",
       "      <td>heat</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>17</td>\n",
       "      <td>al</td>\n",
       "      <td>maqam</td>\n",
       "      <td>arabic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>18</td>\n",
       "      <td>pieces</td>\n",
       "      <td>iraq</td>\n",
       "      <td>shamma</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>19</td>\n",
       "      <td>mentioning</td>\n",
       "      <td>thanks</td>\n",
       "      <td>shamma</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>2</td>\n",
       "      <td>textures</td>\n",
       "      <td>stew</td>\n",
       "      <td>staples</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>20</td>\n",
       "      <td>music</td>\n",
       "      <td>arabs</td>\n",
       "      <td>voice</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>21</td>\n",
       "      <td>music</td>\n",
       "      <td>musician</td>\n",
       "      <td>great</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>22</td>\n",
       "      <td>chose</td>\n",
       "      <td>self</td>\n",
       "      <td>love</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>23</td>\n",
       "      <td>al</td>\n",
       "      <td>songs</td>\n",
       "      <td>like</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>24</td>\n",
       "      <td>songs</td>\n",
       "      <td>playlist</td>\n",
       "      <td>listen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>25</td>\n",
       "      <td>gang</td>\n",
       "      <td>billie</td>\n",
       "      <td>roar</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>26</td>\n",
       "      <td>banks</td>\n",
       "      <td>lana</td>\n",
       "      <td>neighbourhood</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>27</td>\n",
       "      <td>hauser</td>\n",
       "      <td>music</td>\n",
       "      <td>favourite</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>28</td>\n",
       "      <td>dion</td>\n",
       "      <td>celine</td>\n",
       "      <td>lists</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>29</td>\n",
       "      <td>turkish</td>\n",
       "      <td>songs</td>\n",
       "      <td>buble</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>3</td>\n",
       "      <td>fish</td>\n",
       "      <td>iraqi</td>\n",
       "      <td>letting</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>30</td>\n",
       "      <td>sezen</td>\n",
       "      <td>aksu</td>\n",
       "      <td>turkish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>31</td>\n",
       "      <td>series</td>\n",
       "      <td>turkish</td>\n",
       "      <td>language</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>4</td>\n",
       "      <td>kubbah</td>\n",
       "      <td>burghul</td>\n",
       "      <td>mosul</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>5</td>\n",
       "      <td>falafel</td>\n",
       "      <td>sandwich</td>\n",
       "      <td>bread</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>6</td>\n",
       "      <td>khubz</td>\n",
       "      <td>tannour</td>\n",
       "      <td>bread</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>7</td>\n",
       "      <td>searched</td>\n",
       "      <td>mansaf</td>\n",
       "      <td>try</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>8</td>\n",
       "      <td>meat</td>\n",
       "      <td>number</td>\n",
       "      <td>complete</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>iq</td>\n",
       "      <td>9</td>\n",
       "      <td>eid</td>\n",
       "      <td>killecha</td>\n",
       "      <td>fillings</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>1</td>\n",
       "      <td>considering</td>\n",
       "      <td>thank</td>\n",
       "      <td>perfect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>10</td>\n",
       "      <td>song</td>\n",
       "      <td>animation</td>\n",
       "      <td>languages</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>11</td>\n",
       "      <td>artists</td>\n",
       "      <td>listen</td>\n",
       "      <td>hop</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>12</td>\n",
       "      <td>learn</td>\n",
       "      <td>english</td>\n",
       "      <td>songs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>13</td>\n",
       "      <td>artist</td>\n",
       "      <td>pop</td>\n",
       "      <td>double</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>2</td>\n",
       "      <td>mansaf</td>\n",
       "      <td>typically</td>\n",
       "      <td>served</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>3</td>\n",
       "      <td>really</td>\n",
       "      <td>cooking</td>\n",
       "      <td>family</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>4</td>\n",
       "      <td>chocolate</td>\n",
       "      <td>baking</td>\n",
       "      <td>cup</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>5</td>\n",
       "      <td>food</td>\n",
       "      <td>dessert</td>\n",
       "      <td>culture</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>6</td>\n",
       "      <td>broth</td>\n",
       "      <td>noodles</td>\n",
       "      <td>dish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>7</td>\n",
       "      <td>tbl</td>\n",
       "      <td>add</td>\n",
       "      <td>soy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>8</td>\n",
       "      <td>el</td>\n",
       "      <td>songs</td>\n",
       "      <td>enjoy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>foodmusic</td>\n",
       "      <td>us</td>\n",
       "      <td>9</td>\n",
       "      <td>michaels</td>\n",
       "      <td>julia</td>\n",
       "      <td>jp</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Topic State  DocID      Tfidf 1    Tfidf 2        Tfidf 3\n",
       "0   foodmusic    iq      1        dolma        mom           tbsp\n",
       "1   foodmusic    iq     10        kebab    grilled          grill\n",
       "2   foodmusic    iq     11       kibbeh        add           meat\n",
       "3   foodmusic    iq     12     shawarma        tsp        chicken\n",
       "4   foodmusic    iq     13        spicy    totally        pumpkin\n",
       "5   foodmusic    iq     14     potatoes       oven       teaspoon\n",
       "6   foodmusic    iq     15    croissant      cocoa       contains\n",
       "7   foodmusic    iq     16          cup      sugar           heat\n",
       "8   foodmusic    iq     17           al      maqam         arabic\n",
       "9   foodmusic    iq     18       pieces       iraq         shamma\n",
       "10  foodmusic    iq     19   mentioning     thanks         shamma\n",
       "11  foodmusic    iq      2     textures       stew        staples\n",
       "12  foodmusic    iq     20        music      arabs          voice\n",
       "13  foodmusic    iq     21        music   musician          great\n",
       "14  foodmusic    iq     22        chose       self           love\n",
       "15  foodmusic    iq     23           al      songs           like\n",
       "16  foodmusic    iq     24        songs   playlist         listen\n",
       "17  foodmusic    iq     25         gang     billie           roar\n",
       "18  foodmusic    iq     26        banks       lana  neighbourhood\n",
       "19  foodmusic    iq     27       hauser      music      favourite\n",
       "20  foodmusic    iq     28         dion     celine          lists\n",
       "21  foodmusic    iq     29      turkish      songs          buble\n",
       "22  foodmusic    iq      3         fish      iraqi        letting\n",
       "23  foodmusic    iq     30        sezen       aksu        turkish\n",
       "24  foodmusic    iq     31       series    turkish       language\n",
       "25  foodmusic    iq      4       kubbah    burghul          mosul\n",
       "26  foodmusic    iq      5      falafel   sandwich          bread\n",
       "27  foodmusic    iq      6        khubz    tannour          bread\n",
       "28  foodmusic    iq      7     searched     mansaf            try\n",
       "29  foodmusic    iq      8         meat     number       complete\n",
       "30  foodmusic    iq      9          eid   killecha       fillings\n",
       "31  foodmusic    us      1  considering      thank        perfect\n",
       "32  foodmusic    us     10         song  animation      languages\n",
       "33  foodmusic    us     11      artists     listen            hop\n",
       "34  foodmusic    us     12        learn    english          songs\n",
       "35  foodmusic    us     13       artist        pop         double\n",
       "36  foodmusic    us      2       mansaf  typically         served\n",
       "37  foodmusic    us      3       really    cooking         family\n",
       "38  foodmusic    us      4    chocolate     baking            cup\n",
       "39  foodmusic    us      5         food    dessert        culture\n",
       "40  foodmusic    us      6        broth    noodles           dish\n",
       "41  foodmusic    us      7          tbl        add            soy\n",
       "42  foodmusic    us      8           el      songs          enjoy\n",
       "43  foodmusic    us      9     michaels      julia             jp"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_tfidf_terms(docs[docs['TOPIC'] == 'foodmusic'], 3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "185d3c2f",
   "metadata": {},
   "source": [
    "#### Novel words"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "08d2f23e",
   "metadata": {},
   "source": [
    "While most words not found in the WordNet English dictionary are mistyped/misspelled, some are foreign words. These words are interesting in that they indicate cross-cultural communication."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49b800eb",
   "metadata": {},
   "source": [
    "The food/music prompt generated many foreign words."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9b4626ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "novel = novel_words(docs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "6ba91956",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('novel-export-v2.txt', 'w') as f:\n",
    "    for word in novel:\n",
    "        f.write(word)\n",
    "        f.write('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "12f879cb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['kibbeh', 'badna', 'jawza', 'mansaf', 'majid', 'baghdadi', 'nazem', 'ghazali', 'ahlan', 'hammoud', 'zaydan', 'nasir', 'joji', 'dalia', 'haqi', 'khubz', 'mansour', 'ahebek', \"qur'an\", 'saad', '1948', 'sahir', 'hawi', 'dunbug', 'kun', 'rashid', 'qundarchi', 'irmak', 'kahoat', 'armani', 'ajram', 'brockhampton', 'fusul', 'adele', 'dolma', 'moussoula', 'sulic', 'strach', 'hassan', 'samoon', 'adha', 'alkhuder', 'aiko', 'luka', 'kifak', 'quranic', \"a'dhami\", 'siu', 'okhra', 'saif', 'chalghi', 'boz', 'kleichat', 'dumbeg', 'celine', \"a'am\", 'hamid', 'blondie', 'muhur', 'rejab', 'qubbanchi', 'unuturum', 'hantoor', 'maher', 'falastini', 'ghaly', 'najm', 'aleikum', 'spotify', 'husayni', 'lana', 'abdulrahman', 'aksu', 'yusuf', 'ganoush', 'mohem', 'basmati', 'amba', 'kitab', 'thomeyah', 'lughat', 'nished', 'facebook', 'nisaa', 'aksoy', 'mohandis', 'tannour', 'fava', 'eillsh', 'xiu', \"qari'\", 'katy', 'coz', 'eid', 'akhi', 'kubbah', 'khaiwka', 'hussain', 'buble', '1986', 'hokkien', \"beta'alla\", 'pho', 'derya', 'khalid', 'musgraves', 'killecha', 'burghul', 'karanlik', 'khidhr', 'ehab', 'soghayar', 'maqam', 'surah', 'saeed', 'arici', 'hauser', 'aleik', 'fairuz', 'rafet', 'saadoun', 'khobz', 'shammah', 'shaykhli', 'julia', 'nawman', 'tarshey', 'dammi', 'youtube', 'billie', 'nwalee', 'badwah', 'jassmi', \"sa'di\", 'maklouba', 'dion', 'hattan', 'sza', 'sezen', 'nabeel', 'assaf', 'wadaa', 'marra', 'farida', 'jeden', 'yah', 'feek', 'kibbeha', 'tamur', 'cordish', 'kubbeh', 'megan', 'ahmed', 'albamiya', 'taslum', 'eilish', 'toefl', 'alalam', 'naseer', 'ceceli', 'murat', 'cozonac', 'gourgy', 'easer', 'margat', 'salamat', 'omar', 'riqq', 'alsalamu', 'tahina', 'casio', 'cennetten', 'cicek', 'mustafa', 'khiliqtah', 'shawarma', 'tahrir', 'jaber', 'gunay', 'amina', 'elbet', 'zuhayri', 'weeknd', 'naqqarat', 'tawfik', 'kleicha', 'hashim', 'brigham', 'kubba', 'masgouf', 'kadim', 'galeb', 'nawa', 'santur', 'saj', 'zain', 'kut', '2cellos', 'shamma']\n"
     ]
    }
   ],
   "source": [
    "with open('novel-import-v2.txt', 'r') as f:\n",
    "    print(f.read().split('\\n')[:-1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd70e997",
   "metadata": {},
   "source": [
    "### Topic Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1132a6a4",
   "metadata": {},
   "source": [
    "We explore two statistical methods to find topics in documents: latent Dirichlet allocation and non-negative matrix factorization."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8292a5c9",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>language</td>\n",
       "      <td>add</td>\n",
       "      <td>song</td>\n",
       "      <td>woman</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>english</td>\n",
       "      <td>cup</td>\n",
       "      <td>music</td>\n",
       "      <td>gender</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>learning</td>\n",
       "      <td>recipe</td>\n",
       "      <td>love</td>\n",
       "      <td>education</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>learn</td>\n",
       "      <td>bread</td>\n",
       "      <td>favorite</td>\n",
       "      <td>also</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>also</td>\n",
       "      <td>dish</td>\n",
       "      <td>really</td>\n",
       "      <td>learning</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>people</td>\n",
       "      <td>oil</td>\n",
       "      <td>like</td>\n",
       "      <td>student</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>lingua</td>\n",
       "      <td>make</td>\n",
       "      <td>listen</td>\n",
       "      <td>family</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>english_language</td>\n",
       "      <td>put</td>\n",
       "      <td>one</td>\n",
       "      <td>group</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>franca</td>\n",
       "      <td>onion</td>\n",
       "      <td>food</td>\n",
       "      <td>men</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>one</td>\n",
       "      <td>meat</td>\n",
       "      <td>culture</td>\n",
       "      <td>one</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            Topic 1 Topic 2   Topic 3    Topic 4\n",
       "0          language     add      song      woman\n",
       "1           english     cup     music     gender\n",
       "2          learning  recipe      love  education\n",
       "3             learn   bread  favorite       also\n",
       "4              also    dish    really   learning\n",
       "5            people     oil      like    student\n",
       "6            lingua    make    listen     family\n",
       "7  english_language     put       one      group\n",
       "8            franca   onion      food        men\n",
       "9               one    meat   culture        one"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bow, mdl = lda(docs, 4)\n",
    "lda_topics(mdl, bow, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "8960d327",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "      <th>Topic 5</th>\n",
       "      <th>Topic 6</th>\n",
       "      <th>Topic 7</th>\n",
       "      <th>Topic 8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>woman</td>\n",
       "      <td>add</td>\n",
       "      <td>language</td>\n",
       "      <td>learning</td>\n",
       "      <td>song</td>\n",
       "      <td>iraq</td>\n",
       "      <td>make</td>\n",
       "      <td>food</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>gender</td>\n",
       "      <td>cup</td>\n",
       "      <td>english</td>\n",
       "      <td>student</td>\n",
       "      <td>music</td>\n",
       "      <td>school</td>\n",
       "      <td>chicken</td>\n",
       "      <td>culture</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>men</td>\n",
       "      <td>oil</td>\n",
       "      <td>learning</td>\n",
       "      <td>online</td>\n",
       "      <td>favorite</td>\n",
       "      <td>group</td>\n",
       "      <td>bread</td>\n",
       "      <td>dish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>education</td>\n",
       "      <td>onion</td>\n",
       "      <td>learn</td>\n",
       "      <td>also</td>\n",
       "      <td>love</td>\n",
       "      <td>woman</td>\n",
       "      <td>like</td>\n",
       "      <td>different</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>family</td>\n",
       "      <td>meat</td>\n",
       "      <td>one</td>\n",
       "      <td>covid</td>\n",
       "      <td>like</td>\n",
       "      <td>one</td>\n",
       "      <td>reason</td>\n",
       "      <td>part</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>inequality</td>\n",
       "      <td>ingredient</td>\n",
       "      <td>also</td>\n",
       "      <td>time</td>\n",
       "      <td>listen</td>\n",
       "      <td>discussed</td>\n",
       "      <td>heat</td>\n",
       "      <td>country</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>issue</td>\n",
       "      <td>recipe</td>\n",
       "      <td>people</td>\n",
       "      <td>pandemic</td>\n",
       "      <td>iraq</td>\n",
       "      <td>stem</td>\n",
       "      <td>iraqi</td>\n",
       "      <td>recipe</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>society</td>\n",
       "      <td>put</td>\n",
       "      <td>lingua</td>\n",
       "      <td>education</td>\n",
       "      <td>artist</td>\n",
       "      <td>allowed</td>\n",
       "      <td>people</td>\n",
       "      <td>food_important</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>also</td>\n",
       "      <td>salt</td>\n",
       "      <td>english_language</td>\n",
       "      <td>life</td>\n",
       "      <td>really</td>\n",
       "      <td>many</td>\n",
       "      <td>really</td>\n",
       "      <td>mixture</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>discussion</td>\n",
       "      <td>pepper</td>\n",
       "      <td>franca</td>\n",
       "      <td>class</td>\n",
       "      <td>hope</td>\n",
       "      <td>expectation</td>\n",
       "      <td>many</td>\n",
       "      <td>important</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Topic 1     Topic 2           Topic 3    Topic 4   Topic 5      Topic 6  \\\n",
       "0       woman         add          language   learning      song         iraq   \n",
       "1      gender         cup           english    student     music       school   \n",
       "2         men         oil          learning     online  favorite        group   \n",
       "3   education       onion             learn       also      love        woman   \n",
       "4      family        meat               one      covid      like          one   \n",
       "5  inequality  ingredient              also       time    listen    discussed   \n",
       "6       issue      recipe            people   pandemic      iraq         stem   \n",
       "7     society         put            lingua  education    artist      allowed   \n",
       "8        also        salt  english_language       life    really         many   \n",
       "9  discussion      pepper            franca      class      hope  expectation   \n",
       "\n",
       "   Topic 7         Topic 8  \n",
       "0     make            food  \n",
       "1  chicken         culture  \n",
       "2    bread            dish  \n",
       "3     like       different  \n",
       "4   reason            part  \n",
       "5     heat         country  \n",
       "6    iraqi          recipe  \n",
       "7   people  food_important  \n",
       "8   really         mixture  \n",
       "9     many       important  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bow, mdl = lda(docs, 8)\n",
    "lda_topics(mdl, bow, 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19052b62",
   "metadata": {},
   "source": [
    "For latent Dirichlet allocation, we find main groups that correspond to the prompts. Inferring more groups results in many small groups that correspond to specific foods, recipes, and music."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "98a5d6f2",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>searched</td>\n",
       "      <td>english</td>\n",
       "      <td>croissant</td>\n",
       "      <td>songs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mansaf</td>\n",
       "      <td>language</td>\n",
       "      <td>cocoa</td>\n",
       "      <td>love</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>delicious</td>\n",
       "      <td>learning</td>\n",
       "      <td>contains</td>\n",
       "      <td>music</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>try</td>\n",
       "      <td>women</td>\n",
       "      <td>stuffed</td>\n",
       "      <td>turkish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>recipe</td>\n",
       "      <td>also</td>\n",
       "      <td>cozonac</td>\n",
       "      <td>artists</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>dish</td>\n",
       "      <td>lingua</td>\n",
       "      <td>butter</td>\n",
       "      <td>amazing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>really</td>\n",
       "      <td>franca</td>\n",
       "      <td>appetit</td>\n",
       "      <td>really</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>considering</td>\n",
       "      <td>people</td>\n",
       "      <td>bon</td>\n",
       "      <td>listening</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>sometime</td>\n",
       "      <td>group</td>\n",
       "      <td>think</td>\n",
       "      <td>count</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>absolutely</td>\n",
       "      <td>learn</td>\n",
       "      <td>cup</td>\n",
       "      <td>celine</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Topic 1   Topic 2    Topic 3    Topic 4\n",
       "0     searched   english  croissant      songs\n",
       "1       mansaf  language      cocoa       love\n",
       "2    delicious  learning   contains      music\n",
       "3          try     women    stuffed    turkish\n",
       "4       recipe      also    cozonac    artists\n",
       "5         dish    lingua     butter    amazing\n",
       "6       really    franca    appetit     really\n",
       "7  considering    people        bon  listening\n",
       "8     sometime     group      think      count\n",
       "9   absolutely     learn        cup     celine"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorizer, mdl = nmf(docs, 4)\n",
    "nmf_topics(mdl, vectorizer, 4, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "52c3ece4",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "      <th>Topic 5</th>\n",
       "      <th>Topic 6</th>\n",
       "      <th>Topic 7</th>\n",
       "      <th>Topic 8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>searched</td>\n",
       "      <td>english</td>\n",
       "      <td>croissant</td>\n",
       "      <td>songs</td>\n",
       "      <td>sometime</td>\n",
       "      <td>women</td>\n",
       "      <td>mentioning</td>\n",
       "      <td>spicy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mansaf</td>\n",
       "      <td>language</td>\n",
       "      <td>contains</td>\n",
       "      <td>love</td>\n",
       "      <td>considering</td>\n",
       "      <td>gender</td>\n",
       "      <td>shamma</td>\n",
       "      <td>pumpkin</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dish</td>\n",
       "      <td>learning</td>\n",
       "      <td>cocoa</td>\n",
       "      <td>turkish</td>\n",
       "      <td>perfect</td>\n",
       "      <td>education</td>\n",
       "      <td>naseer</td>\n",
       "      <td>totally</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>try</td>\n",
       "      <td>lingua</td>\n",
       "      <td>stuffed</td>\n",
       "      <td>artists</td>\n",
       "      <td>absolutely</td>\n",
       "      <td>men</td>\n",
       "      <td>musician</td>\n",
       "      <td>cooking</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>recipe</td>\n",
       "      <td>franca</td>\n",
       "      <td>cozonac</td>\n",
       "      <td>count</td>\n",
       "      <td>described</td>\n",
       "      <td>inequalities</td>\n",
       "      <td>fan</td>\n",
       "      <td>bread</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>delicious</td>\n",
       "      <td>people</td>\n",
       "      <td>butter</td>\n",
       "      <td>celine</td>\n",
       "      <td>looks</td>\n",
       "      <td>iraq</td>\n",
       "      <td>link</td>\n",
       "      <td>seems</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>really</td>\n",
       "      <td>learn</td>\n",
       "      <td>appetit</td>\n",
       "      <td>dion</td>\n",
       "      <td>sure</td>\n",
       "      <td>family</td>\n",
       "      <td>might</td>\n",
       "      <td>express</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>typically</td>\n",
       "      <td>languages</td>\n",
       "      <td>bon</td>\n",
       "      <td>lists</td>\n",
       "      <td>try</td>\n",
       "      <td>also</td>\n",
       "      <td>music</td>\n",
       "      <td>care</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>served</td>\n",
       "      <td>used</td>\n",
       "      <td>think</td>\n",
       "      <td>listening</td>\n",
       "      <td>recipe</td>\n",
       "      <td>group</td>\n",
       "      <td>amazing</td>\n",
       "      <td>easy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>rice</td>\n",
       "      <td>also</td>\n",
       "      <td>cup</td>\n",
       "      <td>music</td>\n",
       "      <td>delicious</td>\n",
       "      <td>families</td>\n",
       "      <td>thanks</td>\n",
       "      <td>others</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Topic 1    Topic 2    Topic 3    Topic 4      Topic 5       Topic 6  \\\n",
       "0   searched    english  croissant      songs     sometime         women   \n",
       "1     mansaf   language   contains       love  considering        gender   \n",
       "2       dish   learning      cocoa    turkish      perfect     education   \n",
       "3        try     lingua    stuffed    artists   absolutely           men   \n",
       "4     recipe     franca    cozonac      count    described  inequalities   \n",
       "5  delicious     people     butter     celine        looks          iraq   \n",
       "6     really      learn    appetit       dion         sure        family   \n",
       "7  typically  languages        bon      lists          try          also   \n",
       "8     served       used      think  listening       recipe         group   \n",
       "9       rice       also        cup      music    delicious      families   \n",
       "\n",
       "      Topic 7  Topic 8  \n",
       "0  mentioning    spicy  \n",
       "1      shamma  pumpkin  \n",
       "2      naseer  totally  \n",
       "3    musician  cooking  \n",
       "4         fan    bread  \n",
       "5        link    seems  \n",
       "6       might  express  \n",
       "7       music     care  \n",
       "8     amazing     easy  \n",
       "9      thanks   others  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorizer, mdl = nmf(docs, 8)\n",
    "nmf_topics(mdl, vectorizer, 8, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c0325ebc",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "      <th>Topic 5</th>\n",
       "      <th>Topic 6</th>\n",
       "      <th>Topic 7</th>\n",
       "      <th>Topic 8</th>\n",
       "      <th>Topic 9</th>\n",
       "      <th>Topic 10</th>\n",
       "      <th>Topic 11</th>\n",
       "      <th>Topic 12</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>searched</td>\n",
       "      <td>english</td>\n",
       "      <td>croissant</td>\n",
       "      <td>turkish</td>\n",
       "      <td>sometime</td>\n",
       "      <td>women</td>\n",
       "      <td>mentioning</td>\n",
       "      <td>spicy</td>\n",
       "      <td>lists</td>\n",
       "      <td>cup</td>\n",
       "      <td>learning</td>\n",
       "      <td>music</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mansaf</td>\n",
       "      <td>language</td>\n",
       "      <td>cocoa</td>\n",
       "      <td>songs</td>\n",
       "      <td>considering</td>\n",
       "      <td>gender</td>\n",
       "      <td>shamma</td>\n",
       "      <td>pumpkin</td>\n",
       "      <td>count</td>\n",
       "      <td>sugar</td>\n",
       "      <td>pandemic</td>\n",
       "      <td>love</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dish</td>\n",
       "      <td>lingua</td>\n",
       "      <td>contains</td>\n",
       "      <td>sezen</td>\n",
       "      <td>perfect</td>\n",
       "      <td>men</td>\n",
       "      <td>naseer</td>\n",
       "      <td>totally</td>\n",
       "      <td>celine</td>\n",
       "      <td>heat</td>\n",
       "      <td>covid</td>\n",
       "      <td>artist</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>try</td>\n",
       "      <td>franca</td>\n",
       "      <td>stuffed</td>\n",
       "      <td>aksu</td>\n",
       "      <td>absolutely</td>\n",
       "      <td>inequalities</td>\n",
       "      <td>musician</td>\n",
       "      <td>cooking</td>\n",
       "      <td>dion</td>\n",
       "      <td>syrup</td>\n",
       "      <td>students</td>\n",
       "      <td>listen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>recipe</td>\n",
       "      <td>people</td>\n",
       "      <td>cozonac</td>\n",
       "      <td>love</td>\n",
       "      <td>described</td>\n",
       "      <td>education</td>\n",
       "      <td>fan</td>\n",
       "      <td>bread</td>\n",
       "      <td>takes</td>\n",
       "      <td>oil</td>\n",
       "      <td>online</td>\n",
       "      <td>songs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>delicious</td>\n",
       "      <td>learn</td>\n",
       "      <td>butter</td>\n",
       "      <td>hear</td>\n",
       "      <td>looks</td>\n",
       "      <td>iraq</td>\n",
       "      <td>link</td>\n",
       "      <td>seems</td>\n",
       "      <td>artists</td>\n",
       "      <td>milk</td>\n",
       "      <td>education</td>\n",
       "      <td>favorite</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>really</td>\n",
       "      <td>learning</td>\n",
       "      <td>appetit</td>\n",
       "      <td>katy</td>\n",
       "      <td>sure</td>\n",
       "      <td>legal</td>\n",
       "      <td>might</td>\n",
       "      <td>express</td>\n",
       "      <td>days</td>\n",
       "      <td>mixed</td>\n",
       "      <td>talked</td>\n",
       "      <td>khalid</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>typically</td>\n",
       "      <td>languages</td>\n",
       "      <td>bon</td>\n",
       "      <td>adele</td>\n",
       "      <td>try</td>\n",
       "      <td>woman</td>\n",
       "      <td>amazing</td>\n",
       "      <td>care</td>\n",
       "      <td>listening</td>\n",
       "      <td>flour</td>\n",
       "      <td>impact</td>\n",
       "      <td>song</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>served</td>\n",
       "      <td>used</td>\n",
       "      <td>think</td>\n",
       "      <td>perry</td>\n",
       "      <td>recipe</td>\n",
       "      <td>family</td>\n",
       "      <td>thanks</td>\n",
       "      <td>easy</td>\n",
       "      <td>amazing</td>\n",
       "      <td>cardamom</td>\n",
       "      <td>study</td>\n",
       "      <td>artists</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>mom</td>\n",
       "      <td>communicate</td>\n",
       "      <td>eid</td>\n",
       "      <td>michael</td>\n",
       "      <td>delicious</td>\n",
       "      <td>society</td>\n",
       "      <td>everyone</td>\n",
       "      <td>others</td>\n",
       "      <td>songs</td>\n",
       "      <td>water</td>\n",
       "      <td>also</td>\n",
       "      <td>saxe</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Topic 1      Topic 2    Topic 3  Topic 4      Topic 5       Topic 6  \\\n",
       "0   searched      english  croissant  turkish     sometime         women   \n",
       "1     mansaf     language      cocoa    songs  considering        gender   \n",
       "2       dish       lingua   contains    sezen      perfect           men   \n",
       "3        try       franca    stuffed     aksu   absolutely  inequalities   \n",
       "4     recipe       people    cozonac     love    described     education   \n",
       "5  delicious        learn     butter     hear        looks          iraq   \n",
       "6     really     learning    appetit     katy         sure         legal   \n",
       "7  typically    languages        bon    adele          try         woman   \n",
       "8     served         used      think    perry       recipe        family   \n",
       "9        mom  communicate        eid  michael    delicious       society   \n",
       "\n",
       "      Topic 7  Topic 8    Topic 9  Topic 10   Topic 11  Topic 12  \n",
       "0  mentioning    spicy      lists       cup   learning     music  \n",
       "1      shamma  pumpkin      count     sugar   pandemic      love  \n",
       "2      naseer  totally     celine      heat      covid    artist  \n",
       "3    musician  cooking       dion     syrup   students    listen  \n",
       "4         fan    bread      takes       oil     online     songs  \n",
       "5        link    seems    artists      milk  education  favorite  \n",
       "6       might  express       days     mixed     talked    khalid  \n",
       "7     amazing     care  listening     flour     impact      song  \n",
       "8      thanks     easy    amazing  cardamom      study   artists  \n",
       "9    everyone   others      songs     water       also      saxe  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorizer, mdl = nmf(docs, 12)\n",
    "nmf_topics(mdl, vectorizer, 12, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "89340664",
   "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>Topic 1</th>\n",
       "      <th>Topic 2</th>\n",
       "      <th>Topic 3</th>\n",
       "      <th>Topic 4</th>\n",
       "      <th>Topic 5</th>\n",
       "      <th>Topic 6</th>\n",
       "      <th>Topic 7</th>\n",
       "      <th>Topic 8</th>\n",
       "      <th>Topic 9</th>\n",
       "      <th>Topic 10</th>\n",
       "      <th>Topic 11</th>\n",
       "      <th>Topic 12</th>\n",
       "      <th>Topic 13</th>\n",
       "      <th>Topic 14</th>\n",
       "      <th>Topic 15</th>\n",
       "      <th>Topic 16</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>searched</td>\n",
       "      <td>english</td>\n",
       "      <td>croissant</td>\n",
       "      <td>turkish</td>\n",
       "      <td>considering</td>\n",
       "      <td>women</td>\n",
       "      <td>mentioning</td>\n",
       "      <td>spicy</td>\n",
       "      <td>lists</td>\n",
       "      <td>cup</td>\n",
       "      <td>learning</td>\n",
       "      <td>music</td>\n",
       "      <td>iraqi</td>\n",
       "      <td>julia</td>\n",
       "      <td>power</td>\n",
       "      <td>okra</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mansaf</td>\n",
       "      <td>language</td>\n",
       "      <td>cocoa</td>\n",
       "      <td>songs</td>\n",
       "      <td>sometime</td>\n",
       "      <td>gender</td>\n",
       "      <td>shamma</td>\n",
       "      <td>pumpkin</td>\n",
       "      <td>celine</td>\n",
       "      <td>sugar</td>\n",
       "      <td>pandemic</td>\n",
       "      <td>khalid</td>\n",
       "      <td>meat</td>\n",
       "      <td>saxe</td>\n",
       "      <td>still</td>\n",
       "      <td>beef</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>dish</td>\n",
       "      <td>lingua</td>\n",
       "      <td>contains</td>\n",
       "      <td>aksu</td>\n",
       "      <td>looks</td>\n",
       "      <td>men</td>\n",
       "      <td>naseer</td>\n",
       "      <td>totally</td>\n",
       "      <td>count</td>\n",
       "      <td>heat</td>\n",
       "      <td>covid</td>\n",
       "      <td>love</td>\n",
       "      <td>add</td>\n",
       "      <td>michaels</td>\n",
       "      <td>strongly</td>\n",
       "      <td>albamiya</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>try</td>\n",
       "      <td>franca</td>\n",
       "      <td>stuffed</td>\n",
       "      <td>sezen</td>\n",
       "      <td>perfect</td>\n",
       "      <td>iraq</td>\n",
       "      <td>musician</td>\n",
       "      <td>cooking</td>\n",
       "      <td>dion</td>\n",
       "      <td>syrup</td>\n",
       "      <td>students</td>\n",
       "      <td>artists</td>\n",
       "      <td>fish</td>\n",
       "      <td>really</td>\n",
       "      <td>break</td>\n",
       "      <td>textures</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>recipe</td>\n",
       "      <td>people</td>\n",
       "      <td>cozonac</td>\n",
       "      <td>love</td>\n",
       "      <td>absolutely</td>\n",
       "      <td>inequalities</td>\n",
       "      <td>fan</td>\n",
       "      <td>bread</td>\n",
       "      <td>takes</td>\n",
       "      <td>milk</td>\n",
       "      <td>online</td>\n",
       "      <td>listen</td>\n",
       "      <td>kubbah</td>\n",
       "      <td>song</td>\n",
       "      <td>ideally</td>\n",
       "      <td>spiced</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>delicious</td>\n",
       "      <td>learn</td>\n",
       "      <td>butter</td>\n",
       "      <td>hear</td>\n",
       "      <td>described</td>\n",
       "      <td>legal</td>\n",
       "      <td>link</td>\n",
       "      <td>seems</td>\n",
       "      <td>artists</td>\n",
       "      <td>mixed</td>\n",
       "      <td>education</td>\n",
       "      <td>favorite</td>\n",
       "      <td>rice</td>\n",
       "      <td>hey</td>\n",
       "      <td>seemingly</td>\n",
       "      <td>margat</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>really</td>\n",
       "      <td>learning</td>\n",
       "      <td>appetit</td>\n",
       "      <td>perry</td>\n",
       "      <td>sure</td>\n",
       "      <td>education</td>\n",
       "      <td>might</td>\n",
       "      <td>express</td>\n",
       "      <td>days</td>\n",
       "      <td>cardamom</td>\n",
       "      <td>talked</td>\n",
       "      <td>artist</td>\n",
       "      <td>lemon</td>\n",
       "      <td>stupid</td>\n",
       "      <td>linkage</td>\n",
       "      <td>flavors</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>typically</td>\n",
       "      <td>languages</td>\n",
       "      <td>bon</td>\n",
       "      <td>adele</td>\n",
       "      <td>try</td>\n",
       "      <td>woman</td>\n",
       "      <td>amazing</td>\n",
       "      <td>care</td>\n",
       "      <td>listening</td>\n",
       "      <td>flour</td>\n",
       "      <td>impact</td>\n",
       "      <td>songs</td>\n",
       "      <td>tbl</td>\n",
       "      <td>enjoying</td>\n",
       "      <td>address</td>\n",
       "      <td>stew</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>served</td>\n",
       "      <td>used</td>\n",
       "      <td>think</td>\n",
       "      <td>michael</td>\n",
       "      <td>recipe</td>\n",
       "      <td>family</td>\n",
       "      <td>thanks</td>\n",
       "      <td>others</td>\n",
       "      <td>amazing</td>\n",
       "      <td>water</td>\n",
       "      <td>discussion</td>\n",
       "      <td>apple</td>\n",
       "      <td>mosul</td>\n",
       "      <td>probably</td>\n",
       "      <td>masculinity</td>\n",
       "      <td>staples</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>celebrations</td>\n",
       "      <td>communicate</td>\n",
       "      <td>eid</td>\n",
       "      <td>buble</td>\n",
       "      <td>delicious</td>\n",
       "      <td>society</td>\n",
       "      <td>everyone</td>\n",
       "      <td>easy</td>\n",
       "      <td>songs</td>\n",
       "      <td>oil</td>\n",
       "      <td>study</td>\n",
       "      <td>spotify</td>\n",
       "      <td>burghul</td>\n",
       "      <td>talented</td>\n",
       "      <td>unbreakable</td>\n",
       "      <td>classic</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Topic 1      Topic 2    Topic 3  Topic 4      Topic 5       Topic 6  \\\n",
       "0      searched      english  croissant  turkish  considering         women   \n",
       "1        mansaf     language      cocoa    songs     sometime        gender   \n",
       "2          dish       lingua   contains     aksu        looks           men   \n",
       "3           try       franca    stuffed    sezen      perfect          iraq   \n",
       "4        recipe       people    cozonac     love   absolutely  inequalities   \n",
       "5     delicious        learn     butter     hear    described         legal   \n",
       "6        really     learning    appetit    perry         sure     education   \n",
       "7     typically    languages        bon    adele          try         woman   \n",
       "8        served         used      think  michael       recipe        family   \n",
       "9  celebrations  communicate        eid    buble    delicious       society   \n",
       "\n",
       "      Topic 7  Topic 8    Topic 9  Topic 10    Topic 11  Topic 12 Topic 13  \\\n",
       "0  mentioning    spicy      lists       cup    learning     music    iraqi   \n",
       "1      shamma  pumpkin     celine     sugar    pandemic    khalid     meat   \n",
       "2      naseer  totally      count      heat       covid      love      add   \n",
       "3    musician  cooking       dion     syrup    students   artists     fish   \n",
       "4         fan    bread      takes      milk      online    listen   kubbah   \n",
       "5        link    seems    artists     mixed   education  favorite     rice   \n",
       "6       might  express       days  cardamom      talked    artist    lemon   \n",
       "7     amazing     care  listening     flour      impact     songs      tbl   \n",
       "8      thanks   others    amazing     water  discussion     apple    mosul   \n",
       "9    everyone     easy      songs       oil       study   spotify  burghul   \n",
       "\n",
       "   Topic 14     Topic 15  Topic 16  \n",
       "0     julia        power      okra  \n",
       "1      saxe        still      beef  \n",
       "2  michaels     strongly  albamiya  \n",
       "3    really        break  textures  \n",
       "4      song      ideally    spiced  \n",
       "5       hey    seemingly    margat  \n",
       "6    stupid      linkage   flavors  \n",
       "7  enjoying      address      stew  \n",
       "8  probably  masculinity   staples  \n",
       "9  talented  unbreakable   classic  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorizer, mdl = nmf(docs, 16)\n",
    "nmf_topics(mdl, vectorizer, 16, 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5f7f416e",
   "metadata": {},
   "source": [
    "For non-negative matrix factorization, we find the same major groups. These groups are described by more interesting words. NMF fairs better for more groups, but the groups are still not well-separated. Many groups in NMF correspond to food, recipes, and music."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72a0a4c5",
   "metadata": {},
   "source": [
    "### Sentences"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a1a1d40",
   "metadata": {},
   "source": [
    "Words are great, but context is important! Based on key words we notice above, we can search for the sentences where the words reside. We can print these in the notebook, or we can print them to a file."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b5b72ab4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 12\n",
      "Text :  for serving  2 potatoes sliced in ngers and fried lettuce pickled cucumber or any pickles  optional  garlic sauce   thomeyah     1 cup yogurt 1 cup mayonnaise 1 garlic clove finely minced .  directions  cut chicken breast in thin small pieces like shawarma  then mix all the spice in shawarma ingredients with yogurt and vinegar or lemon marinate the chicken for about 10 minutes in refrigerator or until you fry potatoes and make bread .  heat the pan with oil over medium high heat then place the chicken gradually so the pan doesn't cool down leave for several minutes and then stir fry for few minutes until cooked \n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 10\n",
      "Text :  roasting method  mix all ingredients well and knead until the mixture becomes firm .  spread the mixture on the wide skewers of kebabs then grill over charcoal and place the kebabs on the bread .  and pull out the skewers sprinkle with sumac and parsley\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text :  it goes with everything with each meal especially lunch and works for every occasion .  it is composed of the same ingredients as any other bread   flour yeast salt and water .  it has a circular shape is about 8 inches  20 cm  long and 1 inch  3 cm  thick\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 12\n",
      "Text :  serves   5 shawarma wraps prep time   20 minutes cooking time   15 minutes .  ingredients  5 loaves of medium to large saj bread or tortillas chicken shawarma   1 whole boneless skinless chicken breast 1 tsp paprika 1 2 tsp black pepper 1 2 tsp cinnamon 1 2 tsp ginger 1 4 tsp ground cloves 1 4 tsp ground cardamom 1 4 tsp cumin 1 pinch nutmeg 1 pinch galangal 1 tsp salt  or to taste 1 2 tbsp vinegar or lemon juice 1 2 cup yogurt 2 tbsp sunflower oil .  for serving  2 potatoes sliced in ngers and fried lettuce pickled cucumber or any pickles  optional  garlic sauce   thomeyah     1 cup yogurt 1 cup mayonnaise 1 garlic clove finely minced\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 12\n",
      "Text :  for garlic sauce mix yogurt and mayonnaise with garlic .  place a large pan on a very low heat place a bread loaf on it and put it the proper amount of chicken lettuce pickles some of the garlic sauce and fries and start wrapping the loaf as shown in photos above .  once done with all loaves serve and bon appetit\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text :  it's typical to iraq .  khubz or khobz means bread  and tannour refers to the oven in which it is baked this oven could be clay oven which is traditional design aged thousands of years in which the bread is better in flavor than other ovens .  the specificity of the clay tannour is its vertical and cylindrical shape with a base larger than its opening it's about 3 feet high and heats up to extremely high temperatures\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text :  i'd like to talk about the main part of the iraqi meals and i think if it's not found at home there will be a problem .  khubz tannour is a type of flat bread usually a big loaf .  it's typical to iraq\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 12\n",
      "Text :  i really miss those days and i hope that it will come again soon so that we can meet again and also for us to eat shawarma together .  today i will show you how to make it at home with the same delicious taste that you find at any middle eastern restaurant to prepare the bread at home you can find a recipe for saj bread here and if you do not have time you can buy it or use tortillas if not available .  recipe tips  do not overcook the chicken or it will dry\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : us\n",
      "DocID : 5\n",
      "Text :  my favorite romanian dessert that have cherished since i was young was the dessert cozonac .  it is a type of sweetbread lled with sweet walnut paste poppy seed paste or turkish delight and raisins .  it was especially common in our household during a time of celebration like christmas or easter\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : us\n",
      "DocID : 3\n",
      "Text : i'm not sure if this is american or not but pumpkin bread is definitely something we enjoy over here .  i think food is such an important part of culture and can be really unifying\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text :  even though the advent of bakeries made it more common to get khubz already made it is still homemade in many villages or even cities as of today because of its special flavor .  in photos below you can see khubz mei means  water bread  the white one which stems from the absence of oil in its components .  the second type the yellow khubz meat is the same process but you can add the meat onions  green  celery and the important component turmeric  curcuma  to the dough then it's baked\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text : the iraqi bread   khubz tannour .  i'd like to talk about the main part of the iraqi meals and i think if it's not found at home there will be a problem\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 5\n",
      "Text :  the falafel sandwich is beautifully laid out in very organized layers of goodness and served in a very appetizing way and that is the secret of making falafel sandwich so special and very popular .  first it start with the bread any of your favorite bread  pita bread used as a pocket to make a sandwich or flatbread wrapped around the falafel or samoon bread opened from the side to put the falafel in it  second the sauce there are 4 different sauces you can use they are yogurt sauce tahina sauce  oil paste made from ground sesame seeds  hummus paste and baba ganoush eggplant paste .  the third is to organize the falafel inside the bread fourth vegetable salad  cucumber tomatoes green pepper and lettuce  amba  mango pickle condiment  salad and pickled turnips and beets  tarshey  and lastly you drizzle the falafel sandwich with tahini sauce\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 5\n",
      "Text :  i usually like to make my falafel with chickpeas .  in iraq they usually serve the falafel in a samoon bread .  the falafel sandwich is beautifully laid out in very organized layers of goodness and served in a very appetizing way and that is the secret of making falafel sandwich so special and very popular\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 5\n",
      "Text :  first it start with the bread any of your favorite bread  pita bread used as a pocket to make a sandwich or flatbread wrapped around the falafel or samoon bread opened from the side to put the falafel in it  second the sauce there are 4 different sauces you can use they are yogurt sauce tahina sauce  oil paste made from ground sesame seeds  hummus paste and baba ganoush eggplant paste .  the third is to organize the falafel inside the bread fourth vegetable salad  cucumber tomatoes green pepper and lettuce  amba  mango pickle condiment  salad and pickled turnips and beets  tarshey  and lastly you drizzle the falafel sandwich with tahini sauce .  that is how you make your falafel sandwich so heavenly and delicious love it\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 6\n",
      "Text :  historians have refered that khubz belongs back to the sumerian civilization one of the oldest in mesopotamia .  some records have even shown that more than 300 types of bread existed at the time .  nowadays this region covers most of iraq which is why khubz is mostly considered to be an iraqi dish\n",
      "\n",
      "\n",
      "Topic : foodmusic\n",
      "State : iq\n",
      "DocID : 13\n",
      "Text : you're totally right cooking is one of the ways to express care for others .  this bread seems so delicious spicy and i think it's easy to make but i don't have pumpkin now\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# printing\n",
    "for tup in find_word_in_sentences('bread', sent):\n",
    "    print('Topic : ' + tup[0])\n",
    "    print('State : ' + tup[1])\n",
    "    print('DocID : ' + str(tup[2]))\n",
    "    print('Text : ' + tup[5])\n",
    "    print('\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "1218c71b",
   "metadata": {},
   "outputs": [],
   "source": [
    "words = ['bread','song','music','tradition','america','iraq','meat',\n",
    "         'onion','chicken','pepper','salt','love','like','favorite',\n",
    "         'chocolate','recipe','ingredients','turkish','grill',\n",
    "         'sauce','oil','cup','dish'\n",
    "        ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cab094e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# saving\n",
    "import re\n",
    "for word in words:\n",
    "    with open('searches/' + word + '-iq-foodmusic.txt', 'w') as f:\n",
    "        for tup in set(find_word_in_sentences(word, sent[(sent['TOPIC'] == 'foodmusic') & (sent['STATE'] == 'iq')])):\n",
    "            f.write('Topic : ' + tup[0]); f.write('\\n')\n",
    "            f.write('State : ' + tup[1]); f.write('\\n')\n",
    "            f.write('DocID : ' + str(tup[2])); f.write('\\n')\n",
    "            f.write('Text : ' + re.sub(r'\\.', '.\\n', tup[5])); f.write('\\n')\n",
    "            f.write('\\n')\n",
    "    with open('searches/' + word + '-us-foodmusic.txt', 'w') as f:\n",
    "        for tup in set(find_word_in_sentences(word, sent[(sent['TOPIC'] == 'foodmusic') & (sent['STATE'] == 'us')])):\n",
    "            f.write('Topic : ' + tup[0]); f.write('\\n')\n",
    "            f.write('State : ' + tup[1]); f.write('\\n')\n",
    "            f.write('DocID : ' + str(tup[2])); f.write('\\n')\n",
    "            f.write('Text : ' + re.sub(r'\\.', '.\\n', tup[5])); f.write('\\n')\n",
    "            f.write('\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7ed5a540",
   "metadata": {},
   "source": [
    "## End of Notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0015d832",
   "metadata": {},
   "source": [
    "To close, we use the package `coilnlp` to search for sentences containing words that have interesting statistical behavior for the given dataset and research problem. Please review the saved text files to make interpretations and conclusions."
   ]
  }
 ],
 "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}