kym_spotlight_preprocess.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "kym_spotlight_preprocess.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/fedorst/dataeng-project/blob/main/kym_spotlight_preprocess.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DmJvoTN3-tOa"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import json"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IbJ6I_lc-ttt"
      },
      "source": [
        "path = \"/content/drive/My Drive/dataEngProject/kym_spotlight.json\""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8ADJLtB3-veE",
        "outputId": "ca9e70fd-c972-46dd-afdd-4a07792ee59e"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YjjgIZwo-yPB"
      },
      "source": [
        "df = pd.read_json(path, orient=\"index\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "U0mddRYj-5Fj",
        "outputId": "33f457d9-f77b-427d-f0d8-6689e4de30e5"
      },
      "source": [
        "df[\"url\"] = df.index\n",
        "df = df.reset_index()\n",
        "df = df.drop(columns=[\"index\"])\n",
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>@text</th>\n",
              "      <th>@confidence</th>\n",
              "      <th>@support</th>\n",
              "      <th>@types</th>\n",
              "      <th>@sparql</th>\n",
              "      <th>@policy</th>\n",
              "      <th>Resources</th>\n",
              "      <th>url</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>\"This is Relevant To My Interests\" is an expre...</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0</td>\n",
              "      <td></td>\n",
              "      <td></td>\n",
              "      <td>whitelist</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Gene_ex...</td>\n",
              "      <td>https://knowyourmeme.com/memes/this-is-relevan...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ROFLcopter is an Internet slang term used as a...</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0</td>\n",
              "      <td></td>\n",
              "      <td></td>\n",
              "      <td>whitelist</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Interne...</td>\n",
              "      <td>https://knowyourmeme.com/memes/roflcopter</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bitches Don't Know is an exploitable image mac...</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0</td>\n",
              "      <td></td>\n",
              "      <td></td>\n",
              "      <td>whitelist</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Image_m...</td>\n",
              "      <td>https://knowyourmeme.com/memes/bitches-dont-know</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>\"Leave Britney Alone\" is a viral video in whic...</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0</td>\n",
              "      <td></td>\n",
              "      <td></td>\n",
              "      <td>whitelist</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Alone_(...</td>\n",
              "      <td>https://knowyourmeme.com/memes/leave-britney-a...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>O RLY is internet slang for “OH REALLY?” with ...</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0</td>\n",
              "      <td></td>\n",
              "      <td></td>\n",
              "      <td>whitelist</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Interne...</td>\n",
              "      <td>https://knowyourmeme.com/memes/o-rly</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               @text  ...                                                url\n",
              "0  \"This is Relevant To My Interests\" is an expre...  ...  https://knowyourmeme.com/memes/this-is-relevan...\n",
              "1  ROFLcopter is an Internet slang term used as a...  ...          https://knowyourmeme.com/memes/roflcopter\n",
              "2  Bitches Don't Know is an exploitable image mac...  ...   https://knowyourmeme.com/memes/bitches-dont-know\n",
              "3  \"Leave Britney Alone\" is a viral video in whic...  ...  https://knowyourmeme.com/memes/leave-britney-a...\n",
              "4  O RLY is internet slang for “OH REALLY?” with ...  ...               https://knowyourmeme.com/memes/o-rly\n",
              "\n",
              "[5 rows x 8 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "id": "wilsDrJnBtFB",
        "outputId": "ec72b91a-88a2-49fc-a17f-e8cc8d131ab9"
      },
      "source": [
        "df.iloc[0][\"@text\"]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'\"This is Relevant To My Interests\" is an expression used to convey approval and enthusiasm over a given topic. It is most commonly used on media-sharing websites in the form of text-based comments or image macros featuring photographs of concerned-looking animals.'"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Tcs0cA0aCHi3"
      },
      "source": [
        "df.drop(columns=[\"@confidence\", \"@support\", \"@types\", \"@sparql\", \"@policy\"], inplace=True)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 206
        },
        "id": "YPtv63kPBw0I",
        "outputId": "e77618be-5735-44be-b6f4-958cfbcd9c02"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>@text</th>\n",
              "      <th>Resources</th>\n",
              "      <th>url</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>\"This is Relevant To My Interests\" is an expre...</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Gene_ex...</td>\n",
              "      <td>https://knowyourmeme.com/memes/this-is-relevan...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ROFLcopter is an Internet slang term used as a...</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Interne...</td>\n",
              "      <td>https://knowyourmeme.com/memes/roflcopter</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bitches Don't Know is an exploitable image mac...</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Image_m...</td>\n",
              "      <td>https://knowyourmeme.com/memes/bitches-dont-know</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>\"Leave Britney Alone\" is a viral video in whic...</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Alone_(...</td>\n",
              "      <td>https://knowyourmeme.com/memes/leave-britney-a...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>O RLY is internet slang for “OH REALLY?” with ...</td>\n",
              "      <td>[{'@URI': 'http://dbpedia.org/resource/Interne...</td>\n",
              "      <td>https://knowyourmeme.com/memes/o-rly</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               @text  ...                                                url\n",
              "0  \"This is Relevant To My Interests\" is an expre...  ...  https://knowyourmeme.com/memes/this-is-relevan...\n",
              "1  ROFLcopter is an Internet slang term used as a...  ...          https://knowyourmeme.com/memes/roflcopter\n",
              "2  Bitches Don't Know is an exploitable image mac...  ...   https://knowyourmeme.com/memes/bitches-dont-know\n",
              "3  \"Leave Britney Alone\" is a viral video in whic...  ...  https://knowyourmeme.com/memes/leave-britney-a...\n",
              "4  O RLY is internet slang for “OH REALLY?” with ...  ...               https://knowyourmeme.com/memes/o-rly\n",
              "\n",
              "[5 rows x 3 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "eF4RLeQ0CS6C",
        "outputId": "453e148f-07bf-4373-a820-43cabf00c5f8"
      },
      "source": [
        "df.iloc[3][\"@text\"]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'\"Leave Britney Alone\" is a viral video in which YouTube celebrity Chris Crocker cries while pleading with viewers to leave pop star Britney Spears alone.'"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Jxt_BFZ7B3Ao",
        "outputId": "4895c4ba-491a-4bb9-a178-2d265f6919be"
      },
      "source": [
        "df.iloc[3][\"Resources\"]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[{'@URI': 'http://dbpedia.org/resource/Alone_(Heart_song)',\n",
              "  '@offset': '15',\n",
              "  '@percentageOfSecondRank': '0.40237975532092823',\n",
              "  '@similarityScore': '0.5952799492675273',\n",
              "  '@support': '156',\n",
              "  '@surfaceForm': 'Alone',\n",
              "  '@types': 'Wikidata:Q386724,Wikidata:Q2188189,Schema:MusicRecording,Schema:CreativeWork,DBpedia:Work,DBpedia:MusicalWork,DBpedia:Song'},\n",
              " {'@URI': 'http://dbpedia.org/resource/Viral_video',\n",
              "  '@offset': '27',\n",
              "  '@percentageOfSecondRank': '0.0',\n",
              "  '@similarityScore': '1.0',\n",
              "  '@support': '1765',\n",
              "  '@surfaceForm': 'viral video',\n",
              "  '@types': ''},\n",
              " {'@URI': 'http://dbpedia.org/resource/YouTube',\n",
              "  '@offset': '48',\n",
              "  '@percentageOfSecondRank': '7.123706014659695E-6',\n",
              "  '@similarityScore': '0.9999928763420697',\n",
              "  '@support': '49448',\n",
              "  '@surfaceForm': 'YouTube',\n",
              "  '@types': 'Wikidata:Q386724,Schema:WebSite,Schema:CreativeWork,DBpedia:Work,DBpedia:Website'},\n",
              " {'@URI': 'http://dbpedia.org/resource/Chris_Crocker',\n",
              "  '@offset': '66',\n",
              "  '@percentageOfSecondRank': '1.2876352804796738E-18',\n",
              "  '@similarityScore': '1.0',\n",
              "  '@support': '82',\n",
              "  '@surfaceForm': 'Chris Crocker',\n",
              "  '@types': 'Http://xmlns.com/foaf/0.1/Person,Wikidata:Q5,Wikidata:Q24229398,Wikidata:Q215627,DUL:NaturalPerson,DUL:Agent,Schema:Person,DBpedia:Agent,DBpedia:Person'},\n",
              " {'@URI': 'http://dbpedia.org/resource/Pop_music',\n",
              "  '@offset': '123',\n",
              "  '@percentageOfSecondRank': '0.0011222339258988563',\n",
              "  '@similarityScore': '0.9988103494374357',\n",
              "  '@support': '24318',\n",
              "  '@surfaceForm': 'pop',\n",
              "  '@types': 'Wikidata:Q188451,DUL:Concept,DBpedia:TopicalConcept,DBpedia:Genre,DBpedia:MusicGenre'},\n",
              " {'@URI': 'http://dbpedia.org/resource/Britney_Spears',\n",
              "  '@offset': '132',\n",
              "  '@percentageOfSecondRank': '4.3792165769077124E-9',\n",
              "  '@similarityScore': '0.999999995620783',\n",
              "  '@support': '5048',\n",
              "  '@surfaceForm': 'Britney Spears',\n",
              "  '@types': 'Http://xmlns.com/foaf/0.1/Person,Wikidata:Q5,Wikidata:Q24229398,Wikidata:Q215627,DUL:NaturalPerson,DUL:Agent,Schema:Person,DBpedia:Agent,DBpedia:Person'}]"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dRBDV0VwB5C4"
      },
      "source": [
        "# Resources: @types: unique set of matches' DBpedia and Schema as one hot vectors\n",
        "# phrase2vec cross-correlations between memes @text\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EI4fFFRsHSdT",
        "outputId": "a0541701-3886-4448-e905-86f49075c179"
      },
      "source": [
        "def getTypeDict(types):\n",
        "  uniqueTypeList = list(set(j for t in types for j in t.split(\",\")))\n",
        "  tDict = {}\n",
        "  for t in uniqueTypeList:\n",
        "    if \"http:\" in t.lower():\n",
        "      continue \n",
        "    key, value = t.split(\":\")\n",
        "    if key in tDict:\n",
        "      tDict[key].append(value)\n",
        "    else:\n",
        "      tDict[key] = [value]\n",
        "  return tDict\n",
        "\n",
        "def getTypes(row):\n",
        "  if (pd.isna(row)[\"Resources\"]):\n",
        "    return row\n",
        "  types = [x[\"@types\"] for x in row[\"Resources\"] if x[\"@types\"] != \"\"]\n",
        "  uniqueTypeList = list(set(j for t in types for j in t.split(\",\")))\n",
        "  tDict = getTypeDict(types)\n",
        "  if \"DBpedia\" in tDict:\n",
        "    row[\"types_DBpedia\"] = tDict[\"DBpedia\"]\n",
        "  if \"Schema\" in tDict:\n",
        "    row[\"types_Schema\"] = tDict[\"Schema\"]\n",
        "  return row\n",
        "\n",
        "getTypes(df.iloc[3])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "@text            \"Leave Britney Alone\" is a viral video in whic...\n",
              "Resources        [{'@URI': 'http://dbpedia.org/resource/Alone_(...\n",
              "url              https://knowyourmeme.com/memes/leave-britney-a...\n",
              "types_DBpedia    [Song, Agent, MusicGenre, Work, Person, Musica...\n",
              "types_Schema       [MusicRecording, Person, WebSite, CreativeWork]\n",
              "Name: 3, dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8u8g2NcOHRDF"
      },
      "source": [
        "df = df.apply(getTypes, axis=1)\n",
        "df = df.drop(columns=[\"Resources\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 363
        },
        "id": "Hueza_gc9-nV",
        "outputId": "d6419853-813b-4fca-e8fb-272376638384"
      },
      "source": [
        "df.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>@text</th>\n",
              "      <th>types_DBpedia</th>\n",
              "      <th>types_Schema</th>\n",
              "      <th>url</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>\"This is Relevant To My Interests\" is an expre...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>https://knowyourmeme.com/memes/this-is-relevan...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ROFLcopter is an Internet slang term used as a...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>https://knowyourmeme.com/memes/roflcopter</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Bitches Don't Know is an exploitable image mac...</td>\n",
              "      <td>[Software, Work]</td>\n",
              "      <td>[CreativeWork]</td>\n",
              "      <td>https://knowyourmeme.com/memes/bitches-dont-know</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>\"Leave Britney Alone\" is a viral video in whic...</td>\n",
              "      <td>[Song, Agent, MusicGenre, Work, Person, Musica...</td>\n",
              "      <td>[MusicRecording, Person, WebSite, CreativeWork]</td>\n",
              "      <td>https://knowyourmeme.com/memes/leave-britney-a...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>O RLY is internet slang for “OH REALLY?” with ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "      <td>https://knowyourmeme.com/memes/o-rly</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>\"Yatta!\" is a 2001 parody song by a fictional ...</td>\n",
              "      <td>[Place, Song, ArchitecturalStructure, Work, Lo...</td>\n",
              "      <td>[MusicRecording, CreativeWork, Language, Place]</td>\n",
              "      <td>https://knowyourmeme.com/memes/yatta</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>In Soviet Russia, also known as the Russian Re...</td>\n",
              "      <td>[Song, Agent, Comedian, Work, Person, MusicalW...</td>\n",
              "      <td>[MusicRecording, Person, CreativeWork]</td>\n",
              "      <td>https://knowyourmeme.com/memes/in-soviet-russia</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>Domo is a Japanese character often pasted in t...</td>\n",
              "      <td>[Language]</td>\n",
              "      <td>[Language]</td>\n",
              "      <td>https://knowyourmeme.com/memes/domo</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>\"I Like Turtles\" is a phrase first uttered by ...</td>\n",
              "      <td>[Agent, Place, AdministrativeRegion, Location,...</td>\n",
              "      <td>[Event, AdministrativeArea, MusicGroup, Place,...</td>\n",
              "      <td>https://knowyourmeme.com/memes/i-like-turtles</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>Diabeetus is an exploitable soundbite based on...</td>\n",
              "      <td>[Place, Location, Agent, PopulatedPlace, Disea...</td>\n",
              "      <td>[Person, Country, Place]</td>\n",
              "      <td>https://knowyourmeme.com/memes/diabeetus</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                               @text  ...                                                url\n",
              "0  \"This is Relevant To My Interests\" is an expre...  ...  https://knowyourmeme.com/memes/this-is-relevan...\n",
              "1  ROFLcopter is an Internet slang term used as a...  ...          https://knowyourmeme.com/memes/roflcopter\n",
              "2  Bitches Don't Know is an exploitable image mac...  ...   https://knowyourmeme.com/memes/bitches-dont-know\n",
              "3  \"Leave Britney Alone\" is a viral video in whic...  ...  https://knowyourmeme.com/memes/leave-britney-a...\n",
              "4  O RLY is internet slang for “OH REALLY?” with ...  ...               https://knowyourmeme.com/memes/o-rly\n",
              "5  \"Yatta!\" is a 2001 parody song by a fictional ...  ...               https://knowyourmeme.com/memes/yatta\n",
              "6  In Soviet Russia, also known as the Russian Re...  ...    https://knowyourmeme.com/memes/in-soviet-russia\n",
              "7  Domo is a Japanese character often pasted in t...  ...                https://knowyourmeme.com/memes/domo\n",
              "8  \"I Like Turtles\" is a phrase first uttered by ...  ...      https://knowyourmeme.com/memes/i-like-turtles\n",
              "9  Diabeetus is an exploitable soundbite based on...  ...           https://knowyourmeme.com/memes/diabeetus\n",
              "\n",
              "[10 rows x 4 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J9l7jpvXU5CD"
      },
      "source": [
        "from gensim.test.utils import common_texts\n",
        "from gensim.models.doc2vec import Doc2Vec, TaggedDocument\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B0CjOyZVXfmp"
      },
      "source": [
        "documents = [TaggedDocument(doc, [i]) for i, doc in enumerate(common_texts)]\n",
        "model = Doc2Vec(documents, vector_size=8, window=2, min_count=1, workers=4)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NU7k8Mh-XicB"
      },
      "source": [
        "from gensim.test.utils import get_tmpfile\n",
        "\n",
        "fname = get_tmpfile(\"doc2vec_model\")\n",
        "\n",
        "model.save(fname)\n",
        "model = Doc2Vec.load(fname)  # you can continue training with the loaded model!\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "VORURoq7XnLx",
        "outputId": "776a5e09-4b2e-42b6-dba6-8dca1cc8bbe5"
      },
      "source": [
        "model"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<gensim.models.doc2vec.Doc2Vec at 0x7fbfdd1bd110>"
            ]
          },
          "metadata": {},
          "execution_count": 142
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pqGBuRBabHOe"
      },
      "source": [
        "def getVector(row):\n",
        "  words = row[\"@text\"].lower().split(\" \")\n",
        "  words = [''.join(filter(str.isalpha, word)) for word in words]\n",
        "  vector = model.infer_vector(words)\n",
        "  row[\"doc2vec\"] = vector\n",
        "  return row"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XwcZFuo3bSD1"
      },
      "source": [
        "df = df.apply(getVector, axis=1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Gc04zjq8cVCD",
        "outputId": "1907651e-c37f-4daf-d680-45a1ec12acb2"
      },
      "source": [
        "vec1 = df.iloc[3][\"doc2vec\"]\n",
        "vec2 = df.iloc[4][\"doc2vec\"]\n",
        "print(vec1, vec2)\n",
        "dist = np.linalg.norm(vec1-vec2)\n",
        "print(dist)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 0.08039264  0.07167017 -0.02908351 -0.03511304  0.09431612] [ 0.09286018  0.02714852  0.08155558 -0.09250014 -0.03469218]\n",
            "0.1852432\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E_1GsZzSd0Q9"
      },
      "source": [
        "vector_series = df[\"doc2vec\"]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-DJmFW6_eJ8L"
      },
      "source": [
        "def distFun(vec1, vec2):\n",
        "  return np.linalg.norm(vec1-vec2)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3GunYtr1e1a7"
      },
      "source": [
        "from itertools import combinations"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rbdvmymKewPa"
      },
      "source": [
        "combs = list(combinations(df.index.tolist(),2))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wEsH5amye9ae"
      },
      "source": [
        "mat = np.zeros((df.index.size, df.index.size))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rG5qDlDke8bM"
      },
      "source": [
        "for comb in combinations(df.index.tolist(),2):\n",
        "  mat[comb[0], comb[1]] = distFun(df.loc[comb[0],\"doc2vec\"], df.loc[comb[1], \"doc2vec\"])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jNVbgyzWf0CD"
      },
      "source": [
        "import seaborn as sns"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 955
        },
        "id": "HMKCL5cqjtKH",
        "outputId": "86f8d2c2-f834-4a85-b531-7c6245dd963d"
      },
      "source": [
        "plt.rcParams[\"figure.figsize\"] = (16,16)\n",
        "newmat = np.where(mat == 0, mat.max(), mat)\n",
        "sns.heatmap(newmat[:20, :20])\n",
        "plt.xticks(np.arange(0.5, 20.5, 1), [x[:20] for x in df.loc[:20,\"@text\"]], rotation=45, rotation_mode=\"anchor\")\n",
        "plt.yticks(np.arange(0.5, 20.5, 1), [x[:20] for x in df.loc[:20,\"@text\"]], rotation=0)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x1152 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DT_W94DGkGhV",
        "outputId": "e72d1240-4832-4475-f7d6-f2fbfa886694"
      },
      "source": [
        "minidx1, minidx2 = np.unravel_index(np.argmin(newmat), newmat.shape)\n",
        "print(\"Minimum distance:\", newmat.min())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Minimum distance: 0.004370582289993763\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NFKO0Js7nGe9",
        "outputId": "ff5d9983-f61a-438d-b79a-502d0727fba6"
      },
      "source": [
        "maxidx1, maxidx2 = np.unravel_index(np.argmax(mat), mat.shape)\n",
        "print(\"Maximum distance:\", mat.max())"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Maximum distance: 0.3999771773815155\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fwANTrzHmyKS",
        "outputId": "96ea89c5-9ac2-47bb-e830-b5ee74991dc7"
      },
      "source": [
        "print(df.loc[minidx1, \"@text\"], \"\\n\", df.loc[minidx2, \"@text\"])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Life Hacks are tricks, skills or shortcuts that are meant to increase a person's productivity or efficiency in their everyday lives. \n",
            " Jasper Falling is a popular video remixes of the character Jasper from _Steven Universe falling into a crack.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lnTTPtPAnVYC",
        "outputId": "33bcc9c7-79d6-453f-b270-8a5bcbafe2b3"
      },
      "source": [
        "print(df.loc[maxidx1, \"@text\"], \"\\n\", df.loc[maxidx2, \"@text\"])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Beyonce Reading Menu refers to a series of image macros using a photo from singer Beyonce's Instagram of her reading off a restaurant menu to a server. The meme became popular on Twitter, where people captioned what they imagined Beyonce said to the server. \n",
            " \"Ramirez, Do Everything!\" is a catchphrase and image macro series referencing a character from the first-person shooter video game Call of Duty: Modern Warfare 2. Image macros typically feature a centered picture of Sergeant Foley, a non-playable character in the game, with the word \"RAMIREZ!\" on the top line, and an order of some kind on the bottom line.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qI001kQFnbgJ"
      },
      "source": [
        "df.to_parquet(\"/content/drive/My Drive/dataEngProject/kym_spotlight_processed.parquet\")"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}