statistics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "# Step1: Data Preparation"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "f5f5c74096f57045"
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "initial_id",
   "metadata": {
    "collapsed": true,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:22:00.456213700Z",
     "start_time": "2023-08-26T10:21:56.851407600Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     C:\\Users\\15810\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     C:\\Users\\15810\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": "True"
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import necessary libraries\n",
    "import pandas as pd\n",
    "import nltk\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.corpus import stopwords\n",
    "\n",
    "nltk.download('punkt')\n",
    "nltk.download('stopwords')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "outputs": [],
   "source": [
    "# Load the data\n",
    "with open('guangzhou_content.txt', 'r', encoding='utf-8') as file:\n",
    "    articles = file.readlines()"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:22:04.635178400Z",
     "start_time": "2023-08-26T10:22:04.546780700Z"
    }
   },
   "id": "a41048eb192b3541"
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [
    {
     "data": {
      "text/plain": "[['the',\n  'finals',\n  'shanghai',\n  'international',\n  'youth',\n  'piano',\n  'competition',\n  'held',\n  'aug',\n  'drew',\n  'nearly',\n  'contestants',\n  'selected',\n  'tens',\n  'thousands',\n  'children',\n  'adolescents',\n  'administrative',\n  'regions',\n  'organizers',\n  'number',\n  'participants',\n  'taking',\n  'part',\n  'finals',\n  'year',\n  'record',\n  'high',\n  'indicates',\n  'love',\n  'keyboard',\n  'instrument',\n  'still',\n  'going',\n  'strong',\n  'piano',\n  'lessons',\n  'one',\n  'common',\n  'forms',\n  'extracurricular',\n  'activities',\n  'chinese',\n  'parents',\n  'plan',\n  'children',\n  'while',\n  'may',\n  'argue',\n  'many',\n  'children',\n  'choice',\n  'attending',\n  'lessons',\n  'taking',\n  'part',\n  'competition',\n  'personally',\n  'attest',\n  'much',\n  'love',\n  'person',\n  'wei',\n  'changyou',\n  'luohe',\n  'city',\n  'henan',\n  'province',\n  'the',\n  'started',\n  'learning',\n  'piano',\n  'age',\n  'says',\n  'heart',\n  'overflows',\n  'pleasant',\n  'feelings',\n  'listens',\n  'beautiful',\n  'piano',\n  'pieces',\n  'the',\n  'song',\n  'i',\n  'like',\n  'play',\n  'spinning',\n  'song',\n  'albert',\n  'ellmenreich',\n  'german',\n  'composer',\n  'the',\n  'melody',\n  'agile',\n  'light',\n  'flowing',\n  'water',\n  'i',\n  'feel',\n  'like',\n  'magic',\n  'chords',\n  'run',\n  'fingers',\n  'mother',\n  'meng',\n  'hebai',\n  'says',\n  'changyou',\n  'practiced',\n  'piano',\n  'minutes',\n  'day',\n  'first',\n  'started',\n  'learning',\n  'gradually',\n  'became',\n  'hours',\n  'got',\n  'captivated',\n  'changyou',\n  'spends',\n  'average',\n  'two',\n  'hours',\n  'practicing',\n  'musical',\n  'instrument',\n  'daily',\n  'basis',\n  'in',\n  'fact',\n  'even',\n  'practices',\n  'accord',\n  'i',\n  'see',\n  'whether',\n  'feels',\n  'happy',\n  'frustrated',\n  'sit',\n  'piano',\n  'it',\n  'already',\n  'become',\n  'friend',\n  'says',\n  'playing',\n  'also',\n  'helped',\n  'girl',\n  'gain',\n  'mother',\n  'tanyuanjun',\n  'guangzhou',\n  'guangdong',\n  'province',\n  'another',\n  'competitor',\n  'shares',\n  'love',\n  'describes',\n  'playing',\n  'piano',\n  'favorite',\n  'part',\n  'mother',\n  'li',\n  'qiong',\n  'says',\n  'seen',\n  'tanyuanjun',\n  'gone',\n  'using',\n  'piano',\n  'means',\n  'gratification',\n  'providing',\n  'comfort',\n  'years',\n  'ago',\n  'recalls',\n  'tanyuanjun',\n  'fell',\n  'ill',\n  'hospitalized',\n  'several',\n  'days',\n  'during',\n  'period',\n  'asked',\n  'parents',\n  'bring',\n  'piano',\n  'hospital',\n  'ward',\n  'could',\n  'practice',\n  'kill',\n  'time',\n  'what',\n  'transpired',\n  'scene',\n  'moved',\n  'li',\n  'people',\n  'drawn',\n  'music',\n  'formed',\n  'circle',\n  'around',\n  'i',\n  'felt',\n  'deeply',\n  'touched',\n  'moment',\n  'music',\n  'must',\n  'healed',\n  'many',\n  'hospital',\n  'stay',\n  'wang',\n  'xichong',\n  'father',\n  'contestant',\n  'wang',\n  'yukun',\n  'wuhan',\n  'hubei',\n  'province',\n  'objective',\n  'son',\n  'learning',\n  'piano',\n  'mastering',\n  'instrument',\n  'rather',\n  'catharsis',\n  'offers',\n  'i',\n  'learned',\n  'play',\n  'piano',\n  'i',\n  'child',\n  'way',\n  'release',\n  'emotions',\n  'i',\n  'remember',\n  'days',\n  'gaokao',\n  'national',\n  'college',\n  'entrance',\n  'exam',\n  'i',\n  'played',\n  'lot',\n  'i',\n  'needed',\n  'relieve',\n  'pressure',\n  'i',\n  'recalls',\n  'zipan',\n  'started',\n  'learning',\n  'piano',\n  'kindergarten',\n  'shares',\n  'similar',\n  'sentiment',\n  'deng',\n  'currently',\n  'studies',\n  'piano',\n  'rachmaninoff',\n  'conservatory',\n  'music',\n  'russia',\n  'says',\n  'practicing',\n  'exercise',\n  'perseverance',\n  'courage',\n  'every',\n  'time',\n  'open',\n  'new',\n  'sheet',\n  'music',\n  'journey',\n  'begins',\n  'endurance',\n  'pain',\n  'keep',\n  'encouraging',\n  'take',\n  'another',\n  'step',\n  'forward',\n  'but',\n  'mastered',\n  'piece',\n  'start',\n  'dealing',\n  'details',\n  'becomes',\n  'pleasure',\n  'fully',\n  'immerse',\n  'music',\n  'says',\n  'wuhan',\n  'wanghua',\n  'prestigious',\n  'composer',\n  'pianist',\n  'chairman',\n  'judging',\n  'committee',\n  'year',\n  'national',\n  'finals',\n  'says',\n  'could',\n  'clearly',\n  'feel',\n  'children',\n  'love',\n  'piano',\n  'music',\n  'playing',\n  'he',\n  'points',\n  'particularly',\n  'impressed',\n  'contestant',\n  'playing',\n  'piece',\n  'wolfgang',\n  'mozart',\n  'it',\n  'amazing',\n  'see',\n  'child',\n  'today',\n  'able',\n  'express',\n  'western',\n  'classical',\n  'music',\n  'centuries',\n  'ago',\n  'well',\n  'little',\n  'fingers',\n  'the',\n  'child',\n  'really',\n  'immersed',\n  'heart',\n  'hands',\n  'mind',\n  'music',\n  'says',\n  'chu',\n  'presently',\n  'least',\n  'million',\n  'young',\n  'piano',\n  'learners',\n  'country',\n  'making',\n  'popular',\n  'choice',\n  'musical',\n  'instrument',\n  'learn',\n  'china',\n  'since',\n  'beginning',\n  'century',\n  'many',\n  'may',\n  'take',\n  'different',\n  'career',\n  'paths',\n  'grow',\n  'but',\n  'society',\n  'makes',\n  'progresses',\n  'members',\n  'improved',\n  'comprehensive',\n  'qualities',\n  'learning',\n  'music',\n  'one',\n  'way',\n  'achieve',\n  'says'],\n [],\n ['the',\n  'fact',\n  'china',\n  'economy',\n  'continues',\n  'recover',\n  'its',\n  'rebound',\n  'generally',\n  'sound',\n  'track',\n  'china',\n  'remains',\n  'important',\n  'engine',\n  'world',\n  'economic',\n  'first',\n  'half',\n  'year',\n  'china',\n  'gdp',\n  'grew',\n  'percent',\n  'faster',\n  'growth',\n  'rate',\n  'last',\n  'year',\n  'higher',\n  'average',\n  'annual',\n  'growth',\n  'rate',\n  'percent',\n  'three',\n  'years',\n  'three',\n  'percentage',\n  'points',\n  'higher',\n  'growth',\n  'rate',\n  'us',\n  'economy',\n  'according',\n  'international',\n  'monetary',\n  'fund',\n  'world',\n  'economic',\n  'outlook',\n  'updated',\n  'last',\n  'month',\n  'china',\n  'economy',\n  'expected',\n  'grow',\n  'percent',\n  'account',\n  'global',\n  'growth',\n  'economic',\n  'development',\n  'made',\n  'solid',\n  'progress',\n  'bright',\n  'spots',\n  'quality',\n  'quantity',\n  'growth',\n  'consumption',\n  'playing',\n  'significantly',\n  'stronger',\n  'role',\n  'driving',\n  'country',\n  'economic',\n  'growth',\n  'china',\n  'exports',\n  'maintained',\n  'generally',\n  'stable',\n  'global',\n  'market',\n  'share',\n  'first',\n  'seven',\n  'months',\n  'sluggish',\n  'world',\n  'economic',\n  'recovery',\n  'combined',\n  'unabated',\n  'inflation',\n  'financial',\n  'turbulence',\n  'rising',\n  'debt',\n  'pressure',\n  'poses',\n  'significant',\n  'economic',\n  'challenges',\n  'countries',\n  'china',\n  'economic',\n  'recovery',\n  'characterized',\n  'undulating',\n  'progress',\n  'challenges',\n  'expected',\n  'we',\n  'never',\n  'shy',\n  'away',\n  'instead',\n  'respond',\n  'the',\n  'results',\n  'response',\n  'already',\n  'showing',\n  'beginning',\n  'show',\n  'lately',\n  'china',\n  'rolled',\n  'host',\n  'targeted',\n  'substantive',\n  'measures',\n  'aimed',\n  'promoting',\n  'consumption',\n  'boosting',\n  'private',\n  'sector',\n  'attracting',\n  'foreign',\n  'investment',\n  'applauded',\n  'domestic',\n  'foreign',\n  'investors',\n  'certain',\n  'western',\n  'politicians',\n  'media',\n  'exaggerated',\n  'hyped',\n  'current',\n  'difficulties',\n  'china',\n  'economic',\n  'chinese',\n  'economy',\n  'enjoys',\n  'strong',\n  'resilience',\n  'ample',\n  'potential',\n  'robust',\n  'dynamism',\n  'fundamentals',\n  'sustaining',\n  'china',\n  'economic',\n  'growth',\n  'sound',\n  'we',\n  'confidence',\n  'conditions',\n  'ability',\n  'meet',\n  'economic',\n  'social',\n  'development',\n  'targets',\n  'set',\n  'high',\n  'quality',\n  'opening',\n  'broader',\n  'prospects',\n  'china',\n  'economy',\n  'remaining',\n  'source',\n  'strength',\n  'world',\n  'economic',\n  'recovery',\n  'growth'],\n [],\n ['dutch',\n  'dairy',\n  'company',\n  'royal',\n  'frieslandcampina',\n  'said',\n  'remarkable',\n  'business',\n  'performance',\n  'first',\n  'six',\n  'months',\n  'china',\n  'bullish',\n  'growth',\n  'potential',\n  'country',\n  'fueled',\n  'rising',\n  'demand',\n  'infant',\n  'formula',\n  'young',\n  'chinese',\n  'first',\n  'half',\n  'year',\n  'royal',\n  'frieslandcampina',\n  'largest',\n  'dairy',\n  'firm',\n  'netherlands',\n  'achieved',\n  'growth',\n  'china',\n  'in',\n  'latter',\n  'half',\n  'company',\n  'said',\n  'market',\n  'competition',\n  'foreseen',\n  'become',\n  'intense',\n  'still',\n  'expects',\n  'growth',\n  'country',\n  'our',\n  'china',\n  'business',\n  'done',\n  'extremely',\n  'well',\n  'first',\n  'half',\n  'year',\n  'we',\n  'see',\n  'lot',\n  'growth',\n  'opportunities',\n  'our',\n  'future',\n  'much',\n  'linked',\n  'china',\n  'i',\n  'strong',\n  'believer',\n  'country',\n  'said',\n  'jan',\n  'derck',\n  'van',\n  'karnebeek',\n  'ceo',\n  'royal',\n  'frieslandcampina',\n  'we',\n  'continue',\n  'focus',\n  'investing',\n  'country',\n  'including',\n  'focusing',\n  'recruiting',\n  'employees',\n  'business',\n  'growing',\n  'the',\n  'china',\n  'market',\n  'highly',\n  'important',\n  'us',\n  'said',\n  'van',\n  'karnebeek',\n  'visited',\n  'china',\n  'week',\n  'appointed',\n  'ceo',\n  'june',\n  'dutch',\n  'dairy',\n  'cooperative',\n  'manages',\n  'production',\n  'chain',\n  'grass',\n  'glass',\n  'member',\n  'dairy',\n  'farmers',\n  'said',\n  'would',\n  'focus',\n  'quality',\n  'ahead',\n  'confidence',\n  'china',\n  'market',\n  'comes',\n  'chinese',\n  'consumers',\n  'especially',\n  'young',\n  'parents',\n  'increasingly',\n  'looking',\n  'provide',\n  'premium',\n  'products',\n  'experiences',\n  'dairy',\n  'company',\n  'said',\n  'business',\n  'run',\n  'successfully',\n  'major',\n  'chinese',\n  'cities',\n  'beijing',\n  'shanghai',\n  'guangzhou',\n  'guangdong',\n  'province',\n  'it',\n  'plans',\n  'continue',\n  'expand',\n  'footprint',\n  'smaller',\n  'cities',\n  'believes',\n  'plenty',\n  'growth',\n  'opportunities',\n  'ahead',\n  'behind',\n  'us',\n  'in',\n  'long',\n  'term',\n  'i',\n  'positive',\n  'stability',\n  'birth',\n  'rate',\n  'china',\n  'growth',\n  'prospects',\n  'business',\n  'van',\n  'karnebeek',\n  'frieslandcampina',\n  'said',\n  'also',\n  'launched',\n  'products',\n  'europe',\n  'perspective',\n  'helping',\n  'reduce',\n  'carbon',\n  'pandemic',\n  'chinese',\n  'consumers',\n  'shown',\n  'higher',\n  'awareness',\n  'health',\n  'consumed',\n  'dairy',\n  'products',\n  'enhance',\n  'immunity',\n  'nevertheless',\n  'chinese',\n  'residents',\n  'consume',\n  'relatively',\n  'lower',\n  'volumes',\n  'dairy',\n  'products',\n  'compared',\n  'major',\n  'per',\n  'capita',\n  'annual',\n  'dairy',\n  'consumption',\n  'volume',\n  'less',\n  'global',\n  'average',\n  'exists',\n  'significant',\n  'growth',\n  'potential',\n  'according',\n  'china',\n  'dairy',\n  'industry',\n  'association',\n  'we',\n  'focus',\n  'unbalanced',\n  'consumption',\n  'dairy',\n  'products',\n  'urban',\n  'rural',\n  'areas',\n  'china',\n  'continue',\n  'help',\n  'raise',\n  'national',\n  'health',\n  'level',\n  'chinese',\n  'residents',\n  'said',\n  'song',\n  'kungang',\n  'former',\n  'chairman',\n  'china',\n  'dairy',\n  'industry',\n  'royal',\n  'frieslandcampina',\n  'said',\n  'generation',\n  'z',\n  'consumers',\n  'born',\n  'leading',\n  'new',\n  'trend',\n  'dairy',\n  'consumption',\n  'china',\n  'they',\n  'focus',\n  'digestion',\n  'absorption',\n  'dairy',\n  'products',\n  'pay',\n  'particular',\n  'attention',\n  'various',\n  'types',\n  'dairy',\n  'products',\n  'purchasing',\n  'scenarios',\n  'channels']]"
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Tokenize, remove stopwords, and convert to lowercase\n",
    "stop_words = set(stopwords.words('english'))\n",
    "tokenized_articles = [word_tokenize(article) for article in articles]\n",
    "cleaned_articles = [[word.lower() for word in article if word.isalpha() and word not in stop_words] for article in\n",
    "                    tokenized_articles]\n",
    "\n",
    "# Preview first 5 cleaned articles\n",
    "cleaned_articles[:5]"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:22:19.122230600Z",
     "start_time": "2023-08-26T10:22:06.732199900Z"
    }
   },
   "id": "7a63d31ee5fb7ec1"
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Step2: Define Keywords"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "919573ec9d7ad6b8"
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [],
   "source": [
    "female_words = ['woman', 'women', 'female', 'girl', 'girls', 'lady', 'ladies', 'daughter', 'sister', 'mother', 'wife']\n",
    "male_words = ['man', 'men', 'male', 'boy', 'boys', 'gentleman', 'gentlemen', 'son', 'brother', 'father', 'husband']\n",
    "job_words = ['engineer', 'doctor', 'nurse', 'teacher', 'manager', 'worker', 'officer', 'chef', 'driver', 'journalist']"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:22:23.217583400Z",
     "start_time": "2023-08-26T10:22:23.199113600Z"
    }
   },
   "id": "6932809209afa121"
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Step3: Calculate Semantic Distance"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "ac992e00ca9db220"
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [],
   "source": [
    "from gensim.models import KeyedVectors"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:23:30.312178200Z",
     "start_time": "2023-08-26T10:23:29.957287100Z"
    }
   },
   "id": "c4fe7e443979479b"
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "outputs": [
    {
     "data": {
      "text/plain": "({'engineer': 0.04442935284565796,\n  'doctor': 0.2599344023249366,\n  'nurse': 0.33506500043652276,\n  'teacher': 0.26876568320122635,\n  'manager': 0.052751941521736706,\n  'worker': 0.23995340141383084,\n  'officer': 0.10836776155470447,\n  'chef': 0.12439668444990688,\n  'driver': 0.18715927851471034,\n  'journalist': 0.19766382331197913},\n {'engineer': 0.11833999094299295,\n  'doctor': 0.25123278999870474,\n  'nurse': 0.20287424393675543,\n  'teacher': 0.2006505856459791,\n  'manager': 0.06517823641611771,\n  'worker': 0.20333594625646417,\n  'officer': 0.15253144638104874,\n  'chef': 0.13922278142788194,\n  'driver': 0.21868321976878427,\n  'journalist': 0.18804568594152277})"
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = KeyedVectors.load_word2vec_format('GoogleNews-vectors-negative300.bin.gz', binary=True)\n",
    "def average_similarity_word2vec(word, word_list, model):\n",
    "    similarities = []          \n",
    "# Ensure the main word is in the model's vocabulary \n",
    "    if word not in model.key_to_index:       \n",
    "        return 0.0         \n",
    "    for w in word_list:         \n",
    "# Only compute similarity if the secondary word is also in the model's vocabulary   \n",
    "        if w in model.key_to_index:           \n",
    "            similarity = model.similarity(word, w)      \n",
    "            similarities.append(similarity)          \n",
    "    # If no similarities were computed, return 0.0     \n",
    "    if not similarities:         \n",
    "        return 0.0         \n",
    "        # Otherwise, return the average similarity \n",
    "    return sum(similarities) / len(similarities)  \n",
    "        # Recompute the average similarities using the refined function \n",
    "female_similarities = {job: average_similarity_word2vec(job, female_words, model) for job in job_words} \n",
    "male_similarities = {job: average_similarity_word2vec(job, male_words, model) for job in job_words}  \n",
    "female_similarities, male_similarities"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:39:17.201219500Z",
     "start_time": "2023-08-26T10:38:20.077775900Z"
    }
   },
   "id": "f2d5d177193bce00"
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Step 4: Statistical Analysis"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "4979dfd2b18d9aef"
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Female Similarity - Mean: 0.1818487329575212, Median: 0.19241155091334472, Standard Deviation: 0.09720075600942808\n",
      "Male Similarity - Mean: 0.1740094926716252, Median: 0.19434813579375093, Standard Deviation: 0.054948374111159465\n",
      "t-statistic: 0.4171107766924452, p-value: 0.6863708067593893\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd \n",
    "from scipy.stats import ttest_rel  \n",
    "# Assuming df is your DataFrame with the similarity scores \n",
    "df = pd.DataFrame({\n",
    "    'Job': list(female_similarities.keys()),     \n",
    "    'Female Similarity': list(female_similarities.values()),     \n",
    "    'Male Similarity': list(male_similarities.values()) \n",
    "})  \n",
    "# Descriptive Statistics for Female Similarity Scores \n",
    "female_mean = df['Female Similarity'].mean() \n",
    "female_median = df['Female Similarity'].median() \n",
    "female_std = df['Female Similarity'].std()  \n",
    "# Descriptive Statistics for Male Similarity Scores \n",
    "male_mean = df['Male Similarity'].mean() \n",
    "male_median = df['Male Similarity'].median() \n",
    "male_std = df['Male Similarity'].std()  \n",
    "print(f\"Female Similarity - Mean: {female_mean}, Median: {female_median}, Standard Deviation: {female_std}\") \n",
    "print(f\"Male Similarity - Mean: {male_mean}, Median: {male_median}, Standard Deviation: {male_std}\")  \n",
    "# Inferential Statistics: Paired t-test \n",
    "t_stat, p_value = ttest_rel(df['Female Similarity'], df['Male Similarity']) \n",
    "print(f\"t-statistic: {t_stat}, p-value: {p_value}\")"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T10:54:37.799114700Z",
     "start_time": "2023-08-26T10:54:37.736519500Z"
    }
   },
   "id": "37295441bdee107e"
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Step 5: Visualization"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "e964ec0ad4ad03e3"
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 1200x600 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Bar graph visulaization\n",
    "labels = job_words\n",
    "female_vals = [female_similarities[job] for job in labels]\n",
    "male_vals = [male_similarities[job] for job in labels]\n",
    "width = 0.35\n",
    "x = range(len(labels))\n",
    "\n",
    "fig, ax = plt.subplots(figsize = (12,6))\n",
    "rects1 = ax.bar(x, female_value, width, label='Female Words', color = 'lightcoral')\n",
    "rects2 = ax.bar([i + width for i in x], male_value,width, label='Male Words', color = 'lightblue')\n",
    "\n",
    "# Add some text for labels, title and custom x-axis tick labels, etc.\n",
    "ax.set_ylabel('Average Cosine Similarity')\n",
    "ax.set_title('Average Cosine Similarity of Job Words with Female and Male Words')\n",
    "ax.set_xticks([i + width/2 for i in x])\n",
    "ax.set_xticklabels(labels, rotation = 45)\n",
    "ax.legend()\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2023-08-26T11:05:22.726276300Z",
     "start_time": "2023-08-26T11:05:22.396089600Z"
    }
   },
   "id": "21e05060f1806048"
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}