bigram-clustering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bigram Clustering\n",
    "\n",
    "An experimental approach for automated clustering of identical or similar stories. It groups factchecking articles that are about the same story by checking if their headlines contain identical bigrams or word-pairs. Bigrams give promising results as they capture details better than unigrams, for example, they can help sort articles about 'Covid cure' and 'Covid conspiracy' into two distinct groups."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing libraries\n",
    "import os\n",
    "import requests\n",
    "import time\n",
    "from time import sleep\n",
    "from random import uniform\n",
    "import datetime\n",
    "from datetime import date, timezone\n",
    "import csv\n",
    "from pymongo import MongoClient\n",
    "from dotenv import load_dotenv\n",
    "load_dotenv()\n",
    "import os\n",
    "from os import environ\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pprint import pprint\n",
    "import nltk\n",
    "from nltk.corpus import stopwords\n",
    "import spacy\n",
    "import gensim\n",
    "from gensim.utils import simple_preprocess\n",
    "import re\n",
    "from gensim.models.phrases import Phrases, Phraser\n",
    "from nltk import FreqDist\n",
    "import matplotlib\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "import logging\n",
    "import googletrans\n",
    "from googletrans import Translator\n",
    "import json"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get data from factchecking sites MongoDB\n",
    "def initialize_mongo():\n",
    "    mongo_url = \"mongodb+srv://\"+os.environ.get(\"FACTCHECK_DB_USERNAME\")+\":\"+os.environ.get(\"FACTCHECK_DB_PASSWORD\")+\"@tattle-data-fkpmg.mongodb.net/test?retryWrites=true&w=majority&ssl=true&ssl_cert_reqs=CERT_NONE\"   \n",
    "    cli = MongoClient(mongo_url)\n",
    "    db = cli[os.environ.get(\"FACTCHECK_DB_NAME\")]\n",
    "    coll = db[os.environ.get(\"FACTCHECK_DB_COLLECTION\")]\n",
    "    if coll.count_documents({}) > 0:\n",
    "        return coll \n",
    "    else:\n",
    "        print(\"Error accessing Mongo collection\")\n",
    "        sys.exit()\n",
    "        \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "coll = initialize_mongo()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "16722"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coll.count_documents({})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_weekly_data(coll):\n",
    "    pipeline = [\n",
    "        {\"$project\":{\"date_accessed\":\"$date_accessed\", \"date_updated\":\"$date_updated\", 'postID': \"$postID\",'postURL': \"$postURL\",\n",
    "                     \"headline\": \"$headline\", \"docs\": \"$docs\", \"author\": \"$author\", \"domain\": \"$domain\",\n",
    "                     \"date\": {\"$dateFromString\": {\"dateString\": \"$date_updated\"}}}},\n",
    "        {\"$match\": {\"date\": {\"$gte\":datetime.datetime(2020, 9, 28, 0, 0), \"$lt\": datetime.datetime(2020, 10, 5, 0, 0)}}}\n",
    "    ]\n",
    "    \n",
    "    docs = coll.aggregate(pipeline)\n",
    "    return docs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "94\n"
     ]
    }
   ],
   "source": [
    "c=0\n",
    "result=[]\n",
    "docs = get_weekly_data(coll)\n",
    "for doc in docs:\n",
    "    result.append(doc)\n",
    "    c+=1\n",
    "print(c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame(result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_df = df.drop_duplicates(subset=[\"postURL\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "94"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(clean_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    Image Of Deepika Padukone In Solidarity With F...\n",
       "1    Govt Official Breaks Down Over Maharashtra's C...\n",
       "2    Cruise Ship Crossing Corinth Canal In Greece V...\n",
       "Name: headline, dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Snapshot of headlines\n",
    "clean_df[\"headline\"][0:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Defining a function to remove accented characters in the headlines  \n",
    "def remove_accents(sentence):\n",
    "    return \", \".join(simple_preprocess(str(sentence), deacc=True, max_len=100))\n",
    "clean_df[\"unaccented\"] = clean_df[\"headline\"].map(remove_accents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Detecting non-English headlines using regex\n",
    "def detect_lang(text):\n",
    "    pat = re.compile(\"[^\\x00-\\x7F]\") # matches non-English characters\n",
    "    if pat.search(text):\n",
    "        return 0\n",
    "    else:\n",
    "        return 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_df[\"is_english\"] = clean_df[\"unaccented\"].map(detect_lang)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Translating non-English headlines\n",
    "\n",
    "Googletrans is a free library that sends translation requests to the Google Translate API. \n",
    "Random time delays between requests are advised, else Google may (and probably will) block your ip address.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 297 ms, sys: 57.6 ms, total: 354 ms\n",
      "Wall time: 2min 24s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "translator = Translator()\n",
    "clean_df[\"english_headline\"] = \"\"\n",
    "for i, row in clean_df.iterrows():\n",
    "    if row[\"is_english\"] == 0:\n",
    "        clean_df.at[i, \"english_headline\"] = translator.translate(row[\"headline\"]).text\n",
    "        time.sleep(uniform(3,5))\n",
    "    else:\n",
    "        clean_df.at[i, \"english_headline\"] = row[\"headline\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_df.to_csv(\"clean_df.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Combining the headlines\n",
    "# Tokenizing the headlines\n",
    "def sent_to_words(sentences):\n",
    "    for sentence in sentences:\n",
    "        yield (simple_preprocess(str(sentence), deacc = True))\n",
    "        \n",
    "#all_tokens = list(sent_to_words(all_headlines))\n",
    "all_tokens = list(sent_to_words(clean_df[\"english_headline\"]))\n",
    "# Creating stop words list\n",
    "stop_words = stopwords.words(\"english\")\n",
    "# Adding domain words \n",
    "stop_words.extend([\"fake\", \"fact\", \"check\", \"checked\", \"factcheck\", \"news\", \"false\", \n",
    "                   \"falsely\", \"true\", \"truth\", \"viral\", \"video\", \"image\", \"picture\", \n",
    "                   \"photo\", \"claim\", \"claiming\", \"share\", \"clip\", \"misleading\",\"recent\", \"old\",\n",
    "                  \"india\", \"post\", \"medium\"])\n",
    "# Stop word removal\n",
    "data_stopped = [[word for word in doc if word not in stop_words] for doc in all_tokens]\n",
    "# Creating bigrams\n",
    "bigrams_list = []\n",
    "bigram = gensim.models.Phrases(data_stopped, min_count=2)\n",
    "temp=[]\n",
    "for idx in range(len(data_stopped)):\n",
    "    b=[]\n",
    "    for token in bigram[data_stopped[idx]]:\n",
    "        if '_' in token:\n",
    "            # If token is bigram, add it to document\n",
    "            data_stopped[idx].append(token)\n",
    "            bigrams_list.append(token)\n",
    "            b.append(token)\n",
    "    temp.append(b)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "clean_df[\"bigrams\"] = temp\n",
    "bigrams_df = clean_df[clean_df['bigrams'].apply(lambda x: len(x) > 1)]\n",
    "bigrams_df[\"bigrams\"] = bigrams_df[\"bigrams\"].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "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>bigrams</th>\n",
       "      <th>domain</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>['deepika_padukone', 'farmers_protest']</td>\n",
       "      <td>boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>['corinth_canal', 'greece_gujarat']</td>\n",
       "      <td>boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>['health_minister', 'harsh_vardhan']</td>\n",
       "      <td>boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>['yogi_adityanath', 'hathras_victim']</td>\n",
       "      <td>boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>['hathras_incident', 'bjp_leader']</td>\n",
       "      <td>boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>['osama_bin', 'laden_daughter']</td>\n",
       "      <td>hindi.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>['corinth_canal', 'greece_gujarat']</td>\n",
       "      <td>hindi.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>['yogi_adityanath', 'hathras_victim']</td>\n",
       "      <td>hindi.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>['deepika_padukone', 'farmers_protest']</td>\n",
       "      <td>bangla.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>['amit_shah', 'new_york', 'avian_sarcoma']</td>\n",
       "      <td>bangla.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>['health_minister', 'harsh_vardhan']</td>\n",
       "      <td>bangla.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>['hathras_incident', 'bjp_leader']</td>\n",
       "      <td>bangla.boomlive.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>['health_minister', 'harsh_vardhan']</td>\n",
       "      <td>factly.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>['corinth_canal', 'greece_gujarat']</td>\n",
       "      <td>factly.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>['osama_bin', 'laden_daughter']</td>\n",
       "      <td>factly.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>['amit_shah', 'new_york', 'avian_sarcoma']</td>\n",
       "      <td>factly.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>['osama_bin', 'laden_daughter']</td>\n",
       "      <td>newsmobile.in</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>['health_minister', 'harsh_vardhan']</td>\n",
       "      <td>newsmobile.in</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       bigrams              domain\n",
       "0      ['deepika_padukone', 'farmers_protest']         boomlive.in\n",
       "2          ['corinth_canal', 'greece_gujarat']         boomlive.in\n",
       "4         ['health_minister', 'harsh_vardhan']         boomlive.in\n",
       "9        ['yogi_adityanath', 'hathras_victim']         boomlive.in\n",
       "10          ['hathras_incident', 'bjp_leader']         boomlive.in\n",
       "20             ['osama_bin', 'laden_daughter']   hindi.boomlive.in\n",
       "21         ['corinth_canal', 'greece_gujarat']   hindi.boomlive.in\n",
       "24       ['yogi_adityanath', 'hathras_victim']   hindi.boomlive.in\n",
       "29     ['deepika_padukone', 'farmers_protest']  bangla.boomlive.in\n",
       "33  ['amit_shah', 'new_york', 'avian_sarcoma']  bangla.boomlive.in\n",
       "34        ['health_minister', 'harsh_vardhan']  bangla.boomlive.in\n",
       "37          ['hathras_incident', 'bjp_leader']  bangla.boomlive.in\n",
       "53        ['health_minister', 'harsh_vardhan']           factly.in\n",
       "55         ['corinth_canal', 'greece_gujarat']           factly.in\n",
       "56             ['osama_bin', 'laden_daughter']           factly.in\n",
       "60  ['amit_shah', 'new_york', 'avian_sarcoma']           factly.in\n",
       "91             ['osama_bin', 'laden_daughter']       newsmobile.in\n",
       "92        ['health_minister', 'harsh_vardhan']       newsmobile.in"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bigrams_df[bigrams_df.duplicated(subset=[\"bigrams\"],  keep=False)][[\"bigrams\",\"domain\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "grouped = bigrams_df.groupby(\"bigrams\").count().reset_index()\n",
    "sorted = grouped.sort_values(\"_id\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "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>bigrams</th>\n",
       "      <th>_id</th>\n",
       "      <th>date_accessed</th>\n",
       "      <th>date_updated</th>\n",
       "      <th>postID</th>\n",
       "      <th>postURL</th>\n",
       "      <th>headline</th>\n",
       "      <th>docs</th>\n",
       "      <th>author</th>\n",
       "      <th>domain</th>\n",
       "      <th>date</th>\n",
       "      <th>unaccented</th>\n",
       "      <th>is_english</th>\n",
       "      <th>english_headline</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>['health_minister', 'harsh_vardhan']</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>['corinth_canal', 'greece_gujarat']</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>['osama_bin', 'laden_daughter']</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>['amit_shah', 'new_york', 'avian_sarcoma']</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>['deepika_padukone', 'farmers_protest']</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>['hathras_incident', 'bjp_leader']</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>['yogi_adityanath', 'hathras_victim']</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>['amit_shah', 'avian_sarcoma']</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>['amit_shah', 'new_york']</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                      bigrams  _id  date_accessed  \\\n",
       "6        ['health_minister', 'harsh_vardhan']    4              4   \n",
       "3         ['corinth_canal', 'greece_gujarat']    3              3   \n",
       "7             ['osama_bin', 'laden_daughter']    3              3   \n",
       "1  ['amit_shah', 'new_york', 'avian_sarcoma']    2              2   \n",
       "4     ['deepika_padukone', 'farmers_protest']    2              2   \n",
       "5          ['hathras_incident', 'bjp_leader']    2              2   \n",
       "8       ['yogi_adityanath', 'hathras_victim']    2              2   \n",
       "0              ['amit_shah', 'avian_sarcoma']    1              1   \n",
       "2                   ['amit_shah', 'new_york']    1              1   \n",
       "\n",
       "   date_updated  postID  postURL  headline  docs  author  domain  date  \\\n",
       "6             4       4        4         4     4       4       4     4   \n",
       "3             3       3        3         3     3       3       3     3   \n",
       "7             3       3        3         3     3       3       3     3   \n",
       "1             2       2        2         2     2       2       2     2   \n",
       "4             2       2        2         2     2       2       2     2   \n",
       "5             2       2        2         2     2       2       2     2   \n",
       "8             2       2        2         2     2       2       2     2   \n",
       "0             1       1        1         1     1       1       1     1   \n",
       "2             1       1        1         1     1       1       1     1   \n",
       "\n",
       "   unaccented  is_english  english_headline  \n",
       "6           4           4                 4  \n",
       "3           3           3                 3  \n",
       "7           3           3                 3  \n",
       "1           2           2                 2  \n",
       "4           2           2                 2  \n",
       "5           2           2                 2  \n",
       "8           2           2                 2  \n",
       "0           1           1                 1  \n",
       "2           1           1                 1  "
      ]
     },
     "execution_count": 152,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4     2016 Video Falsely Shared As Health Minister Harsh Vardhan Assaulted                                    \n",
      "34    স্বাস্থ্যমন্ত্রী হর্ষ বর্ধনকে নিগৃহ? মিথ্যে দাবিতে ছড়াল ২০১৬'র ভিডিও                                   \n",
      "53     Assaulted BJP party worker in West Bengal is being portrayed as the Union Health Minister Harsh Vardhan\n",
      "92    Fact Check: The man being assaulted by a mob in the video is NOT Health Minister Dr Harsh Vardhan       \n",
      "Name: headline, dtype: object\n",
      "4     https://www.boomlive.in/fake-news/2016-video-falsely-shared-as-health-minister-harsh-vardhan-assaulted-9971                            \n",
      "34    https://bangla.boomlive.in/fake-news/2016-video-falsely-shared-as-health-minister-harsh-vardhan-assaulted-10001                        \n",
      "53    https://factly.in/assaulted-bjp-party-worker-in-west-bengal-is-being-portrayed-as-the-union-health-minister-harsh-vardhan/             \n",
      "92    http://newsmobile.in/articles/2020/09/30/fact-check-the-man-being-assaulted-by-a-mob-in-the-video-is-not-health-minister-harsh-vardhan/\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "2     Cruise Ship Crossing Corinth Canal In Greece Viral As Gujarat                  \n",
      "21    ग्रीस की कोरिंथ नहर से गुज़रता क्रूज़ शिप का वीडियो गुजरात बताकर वायरल           \n",
      "55     This cruising ship video is from the Corinth Canal in Greece, not from Gujarat\n",
      "Name: headline, dtype: object\n",
      "2     https://www.boomlive.in/fake-news/cruise-ship-crossing-corinth-canal-in-greece-viral-as-gujarat-9954\n",
      "21    https://hindi.boomlive.in/fake-news/video-from-greece-viral-as-cruise-ship-service-in-gujarat-9993  \n",
      "55    https://factly.in/this-cruising-ship-video-is-from-the-corinth-canal-in-greece-not-from-gujarat/    \n",
      "Name: postURL, dtype: object\n",
      "\n",
      "20    क्या ओसामा बिन लादेन की बेटी ज़ोया ने एक भोजपुरी गायक से शादी कर ली है?                     \n",
      "56     Image of Pak model is peddled as Osama bin Laden’s daughter getting married to a Hindu.   \n",
      "91    Fact Check: Is Osama Bin Laden’s daughter Zoya getting married to a Hindu? Here’s the truth\n",
      "Name: headline, dtype: object\n",
      "20    https://hindi.boomlive.in/fact-check/photo-of-bhojpuri-singer-peddled-as-osama-bin-ladens-son-in-law-9983                        \n",
      "56    https://factly.in/image-of-pak-model-is-peddled-as-osama-bin-ladens-daughter-getting-married-to-a-hindu/                         \n",
      "91    http://newsmobile.in/articles/2020/09/30/fact-check-is-osama-bin-ladens-daughter-zoya-getting-married-to-a-hindu-heres-the-truth/\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "33    অ্যাভিয়ান সারকোমা সারাতে অমিত শাহ নিউইয়র্কে, ভাইরাল হল মিথ্যে বার্তা\n",
      "60     Amit Shah did not fly to New York for Avian Sarcoma treatment       \n",
      "Name: headline, dtype: object\n",
      "33    https://bangla.boomlive.in/fake-news/viral-message-falsely-claims-amit-shah-is-in-ny-to-treat-avian-sarcoma-9994\n",
      "60    https://factly.in/amit-shah-did-not-fly-to-new-york-to-treat-avian-sarcoma/                                     \n",
      "Name: postURL, dtype: object\n",
      "\n",
      "0     Image Of Deepika Padukone In Solidarity With Farmers Protest Is Morphed\n",
      "29    কৃষক বিক্ষোভের সমর্থনে দীপিকা পাড়ুকোনের ভাইরাল ছবিটি সম্পাদিত         \n",
      "Name: headline, dtype: object\n",
      "0     https://www.boomlive.in/fake-news/image-of-deepika-padukone-in-solidarity-with-farmers-protest-is-morphed-9944   \n",
      "29    https://bangla.boomlive.in/fake-news/image-of-deepika-padukone-in-solidarity-with-farmers-protest-is-morphed-9950\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "10    Hathras Incident: Photo Of UP BJP Leader Peddled As Accused's Father\n",
      "37    হাথরস ঘটনা: অভিযুক্তের বাবা বলে ভাইরাল ইউপির বিজেপি নেতার ছবি       \n",
      "Name: headline, dtype: object\n",
      "10    https://www.boomlive.in/fake-news/hathras-incident-photo-of-up-bjp-leader-peddled-as-accuseds-father-10021   \n",
      "37    https://bangla.boomlive.in/fake-news/hathras-incident-photo-of-up-bjp-leader-peddled-as-accuseds-father-10026\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "9     Photo Of Yogi Adityanath Watching Hathras Victim's Cremation Is Fake                 \n",
      "24    हाथरस पीड़िता के अंतिम संस्कार का लाइव प्रसारण देखते योगी आदित्यनाथ की तस्वीर फ़र्ज़ी है\n",
      "Name: headline, dtype: object\n",
      "9     https://www.boomlive.in/fake-news/photo-of-yogi-adityanath-watching-hathras-victims-cremation-is-fake-10012  \n",
      "24    https://hindi.boomlive.in/fake-news/photo-of-yogi-adityanath-watching-hathras-victims-cremation-is-fake-10016\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "5    Viral Message Falsely Claims Amit Shah Is In NY To Treat Avian Sarcoma\n",
      "Name: headline, dtype: object\n",
      "5    https://www.boomlive.in/fake-news/viral-message-falsely-claims-amit-shah-is-in-ny-to-treat-avian-sarcoma-9981\n",
      "Name: postURL, dtype: object\n",
      "\n",
      "93    Fact Check: Has Home Minister Amit Shah been flown to New York for treatment? Here’s the truth\n",
      "Name: headline, dtype: object\n",
      "93    http://newsmobile.in/articles/2020/09/30/fact-check-has-home-minister-amit-shah-been-flown-to-new-york-for-treatment-heres-the-truth/\n",
      "Name: postURL, dtype: object\n",
      "\n"
     ]
    }
   ],
   "source": [
    "for data in sorted[\"bigrams\"]:\n",
    "    print(bigrams_df[bigrams_df[\"bigrams\"]==data][\"headline\"])\n",
    "    print(bigrams_df[bigrams_df[\"bigrams\"]==data][\"postURL\"])\n",
    "    print(\"\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "fd = FreqDist()\n",
    "for b in bigrams_list:\n",
    "    fd[b] += 1\n",
    "    \n",
    "fd_sorted={k: v for k, v in sorted(fd.items(), key=lambda item: item[1], reverse=True)}\n",
    "fd_sorted=dict(list(fd_sorted.items())[0: 100])  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'hathras_victim': 6,\n",
       " 'deepika_padukone': 5,\n",
       " 'yogi_adityanath': 5,\n",
       " 'farmers_protest': 4,\n",
       " 'corinth_canal': 4,\n",
       " 'health_minister': 4,\n",
       " 'harsh_vardhan': 4,\n",
       " 'amit_shah': 4,\n",
       " 'bjp_leader': 4,\n",
       " 'osama_bin': 4,\n",
       " 'greece_gujarat': 3,\n",
       " 'avian_sarcoma': 3,\n",
       " 'hathras_incident': 3,\n",
       " 'drug_company': 3,\n",
       " 'laden_daughter': 3,\n",
       " 'new_york': 3,\n",
       " 'minister_harsh': 1}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fd_sorted"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15, 10))\n",
    "plt.suptitle(\"Top bigrams in story headlines\", fontsize= 15)\n",
    "plt.title(\"source: Tattle archive\", fontdict = {'fontsize': 12})\n",
    "plt.xlabel(xlabel = \"Bigram\", fontdict = {'fontsize': 13})\n",
    "plt.ylabel(ylabel = \"Count\", fontdict = {'fontsize': 13})\n",
    "plt.xticks(rotation=90, size =12)\n",
    "plt.yticks(np.arange(0, 10, ))\n",
    "plt.plot(list(fd_sorted.keys()), list(fd_sorted.values()))\n",
    "plt.grid()\n",
    "plt.tight_layout(pad=4)\n",
    "plt.savefig(\"bigram_plot.png\", bbox_inches='tight')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}