CaseStudy3 (1).ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VDacajl0Da76"
      },
      "source": [
        "#  Spam Prediction with Clustering\n",
        "## Danny Chang\n",
        "\n",
        "In an era of constant digital communication, the relentless influx of messages and emails can be overwhelming. Among these, a significant portion comprises unwanted and potentially harmful spam messages, which can disrupt productivity and pose security risks. To combat this issue, we embark on a Data Science project aimed at developing a robust spam classification system.\n",
        "\n",
        "Our goal is to create an intelligent algorithm that can automatically differentiate between legitimate messages and spam, providing users with a clutter-free and secure communication experience. By leveraging the power of machine learning and data analysis, we intend to build a predictive model capable of classifying messages and emails as either \"Spam\" or \"Not Spam\" with a high degree of accuracy.\n",
        "\n",
        "This project will entail various stages, including data collection, preprocessing, feature engineering, model selection, and evaluation. We will draw upon a diverse dataset of messages and emails, encompassing a wide range of characteristics, to train and fine-tune our classification model. Throughout the process, we will explore advanced techniques in natural language processing (NLP) and machine learning to enhance our model's performance and adaptability.\n",
        "\n",
        "The successful completion of this project will not only help individuals manage their digital communications more effectively but also have broader applications in email filtering, cybersecurity, and information management. By mitigating the impact of spam, we aim to contribute to a safer and more efficient digital communication environment."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "mSa-gsA_iagt"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import os\n",
        "import email\n",
        "from collections import Counter\n",
        "import shutil"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "From rssfeeds@jmason.org  Mon Sep 30 13:43:46 2002\n",
            "Return-Path: <rssfeeds@example.com>\n",
            "Delivered-To: yyyy@localhost.example.com\n",
            "Received: from localhost (jalapeno [127.0.0.1])\n",
            "\tby jmason.org (Postfix) with ESMTP id AE79816F16\n",
            "\tfor <jm@localhost>; Mon, 30 Sep 2002 13:43:46 +0100 (IST)\n",
            "Received: from jalapeno [127.0.0.1]\n",
            "\tby localhost with IMAP (fetchmail-5.9.0)\n",
            "\tfor jm@localhost (single-drop); Mon, 30 Sep 2002 13:43:46 +0100 (IST)\n",
            "Received: from dogma.slashnull.org (localhost [127.0.0.1]) by\n",
            "    dogma.slashnull.org (8.11.6/8.11.6) with ESMTP id g8U81fg21359 for\n",
            "    <jm@jmason.org>; Mon, 30 Sep 2002 09:01:41 +0100\n",
            "Message-Id: <200209300801.g8U81fg21359@dogma.slashnull.org>\n",
            "To: yyyy@example.com\n",
            "From: gamasutra <rssfeeds@example.com>\n",
            "Subject: Priceless Rubens works stolen in raid on mansion\n",
            "Date: Mon, 30 Sep 2002 08:01:41 -0000\n",
            "Content-Type: text/plain; encoding=utf-8\n",
            "Lines: 6\n",
            "X-Spam-Status: No, hits=-527.4 required=5.0\n",
            "\ttests=AWL,DATE_IN_PAST_03_06,T_URI_COUNT_0_1\n",
            "\tversion=2.50-cvs\n",
            "X-Spam-Level: \n",
            "\n",
            "URL: http://www.newsisfree.com/click/-1,8381145,215/\n",
            "Date: 2002-09-30T03:04:58+01:00\n",
            "\n",
            "*Arts:* Fourth art raid on philanthropist's home once targeted by the IRA and \n",
            "Dublin gangster Martin Cahill.\n",
            "\n",
            "\n",
            "\n"
          ]
        }
      ],
      "source": [
        "x = os.listdir(\"data/easy_ham\")\n",
        "\n",
        "with open(os.path.join(\"data/easy_ham\",x[0]), \"r\") as file_handler:\n",
        "    msg = file_handler.read()\n",
        "    print(msg)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [],
      "source": [
        "file_name = []\n",
        "label = []\n",
        "\n",
        "# Retriving the data\n",
        "for root,dirs,files in os.walk(\"data/\"):\n",
        "    for f in files:\n",
        "        if \"spam\" in root:\n",
        "            label.append(1)\n",
        "        else:\n",
        "            label.append(0)\n",
        "        file_name.append(os.path.join(root,f))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "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>Message</th>\n",
              "      <th>Target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>data/spam/00249.5f45607c1bffe89f60ba1ec9f878039a</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>data/spam/0355.94ebf637e4bd3db8a81c8ce68ecf681d</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>data/spam/0395.bb934e8b4c39d5eab38f828a26f760b4</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>data/spam/0485.9021367278833179285091e5201f5854</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>data/spam/00373.ebe8670ac56b04125c25100a36ab0510</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                            Message  Target\n",
              "0  data/spam/00249.5f45607c1bffe89f60ba1ec9f878039a       1\n",
              "1   data/spam/0355.94ebf637e4bd3db8a81c8ce68ecf681d       1\n",
              "2   data/spam/0395.bb934e8b4c39d5eab38f828a26f760b4       1\n",
              "3   data/spam/0485.9021367278833179285091e5201f5854       1\n",
              "4  data/spam/00373.ebe8670ac56b04125c25100a36ab0510       1"
            ]
          },
          "execution_count": 6,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "data = pd.DataFrame({\"Message\":file_name,\"Target\":label})\n",
        "data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "sns.countplot(data = data, x = \"Target\")\n",
        "plt.title(\"Spam Count Plot\");"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Data Preprocessing and Cleaning\n",
        "\n",
        "We received a collection of over 9,000 emails for classification purposes. Before we dive into the classification task, it's crucial to understand the nature of our dataset. Initially, the dataset was delivered in a zip file, which comprised five folders containing numerous documents, each holding various emails. \n",
        "\n",
        "Upon preliminary examination of some of the files, it became clear that the dataset had certain characteristics that needed addressing. Aside from the main content of the emails, there were instances where emails were encapsulated within HTML tags. Additionally, the presence of special characters and email metadata were also observed. These elements underscore the importance of parsing the data to extract relevant content, which is essential for accurate classification. \n",
        "\n",
        "The following sections will detail our approach to navigate through and parse the emails in our dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aQeD77T6vIv-",
        "outputId": "33e012b0-22b7-4aeb-f755-8b773e598b00"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "content type content transfer encoding\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "[nltk_data] Error loading punkt: <urlopen error [SSL:\n",
            "[nltk_data]     CERTIFICATE_VERIFY_FAILED] certificate verify failed:\n",
            "[nltk_data]     unable to get local issuer certificate (_ssl.c:1002)>\n",
            "[nltk_data] Error loading stopwords: <urlopen error [SSL:\n",
            "[nltk_data]     CERTIFICATE_VERIFY_FAILED] certificate verify failed:\n",
            "[nltk_data]     unable to get local issuer certificate (_ssl.c:1002)>\n"
          ]
        }
      ],
      "source": [
        "import re\n",
        "from bs4 import BeautifulSoup\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.tokenize import word_tokenize\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')\n",
        "\n",
        "# function to clean data\n",
        "def clean_text(text):\n",
        "\n",
        "    text = str(text)\n",
        "\n",
        "    # Remove HTML tags\n",
        "    text = BeautifulSoup(text, \"html.parser\").get_text()\n",
        "\n",
        "\n",
        "    # Remove special characters and convert to lowercase\n",
        "    text = re.sub(r'[^a-zA-Z0-9]', ' ', text)\n",
        "    #words = re.findall(r'\\w+', text)\n",
        "    #text= ' '.join(words)\n",
        "    text = text.lower()\n",
        "\n",
        "    # Tokenization and stopword removal\n",
        "    tokens = word_tokenize(text)\n",
        "    tokens = [word for word in tokens if word not in stopwords.words(\"english\")]\n",
        "\n",
        "    return \" \".join(tokens)\n",
        "\n",
        "# Example usage:\n",
        "text1 = \"Sample <b>HTML</b> text with special characters & symbols!\"\n",
        "text2 =\"Martin A posted:\\nTassos Papadopoulos, the Greek sculptor behind the plan, judged that the\\n limestone of Mount Kerdylio, 70 miles east of Salonika and not far from the\\n Mount Athos monastic community, was ideal for the patriotic sculpture. \\n \\n As well as Alexander's granite features, 240 ft high and 170 ft wide, a\\n museum, a restored amphitheatre and car park for admiring crowds are\\nplanned\\n---------------------\\nSo is this mountain limestone or granite?\\nIf it's limestone, it'll weather pretty fast.\\n\\n------------------------ Yahoo! Groups Sponsor ---------------------~-->\\n4 DVDs Free +s&p Join Now\\nhttp://us.click.yahoo.com/pt6YBB/NXiEAA/mG3HAA/7gSolB/TM\\n---------------------------------------------------------------------~->\\n\\nTo unsubscribe from this group, send an email to:\\nforteana-unsubscribe@egroups.com\\n\\n \\n\\nYour use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ \\n\\n\\n\\n\"\n",
        "text3 ='    Date:        Wed, 21 Aug 2002 10:54:46 -0500\\n    From:        Chris Garrigues <cwg-dated-1030377287.06fa6d@DeepEddy.Com>\\n    Message-ID:  <1029945287.4797.TMDA@deepeddy.vircio.com>\\n\\n\\n  | I can\\'t reproduce this error.\\n\\nFor me it is very repeatable... (like every time, without fail).\\n\\nThis is the debug log of the pick happening ...\\n\\n18:19:03 Pick_It {exec pick +inbox -list -lbrace -lbrace -subject ftp -rbrace -rbrace} {4852-4852 -sequence mercury}\\n18:19:03 exec pick +inbox -list -lbrace -lbrace -subject ftp -rbrace -rbrace 4852-4852 -sequence mercury\\n18:19:04 Ftoc_PickMsgs {{1 hit}}\\n18:19:04 Marking 1 hits\\n18:19:04 tkerror: syntax error in expression \"int ...\\n\\nNote, if I run the pick command by hand ...\\n\\ndelta$ pick +inbox -list -lbrace -lbrace -subject ftp -rbrace -rbrace  4852-4852 -sequence mercury\\n1 hit\\n\\nThat\\'s where the \"1 hit\" comes from (obviously).  The version of nmh I\\'m\\nusing is ...\\n\\ndelta$ pick -version\\npick -- nmh-1.0.4 [compiled on fuchsia.cs.mu.OZ.AU at Sun Mar 17 14:55:56 ICT 2002]\\n\\nAnd the relevant part of my .mh_profile ...\\n\\ndelta$ mhparam pick\\n-seq sel -list\\n\\n\\nSince the pick command works, the sequence (actually, both of them, the\\none that\\'s explicit on the command line, from the search popup, and the\\none that comes from .mh_profile) do get created.\\n\\nkre\\n\\nps: this is still using the version of the code form a day ago, I haven\\'t\\nbeen able to reach the cvs repository today (local routing issue I think).\\n\\n\\n\\n_______________________________________________\\nExmh-workers mailing list\\nExmh-workers@redhat.com\\nhttps://listman.redhat.com/mailman/listinfo/exmh-workers\\n\\n'\n",
        "text4=\"[[Content-Type, Content-Transfer-Encoding]]\t\"\n",
        "print(clean_text(text4))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 300
        },
        "id": "In5aCqdX1TuH",
        "outputId": "475c5a9a-2ee8-4945-c446-06c735fd7d0e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "data/spam 0343.0630afbe4ee1ffd0db0ffb81c6de98de\n",
            "______________________\n"
          ]
        }
      ],
      "source": [
        "types = Counter()\n",
        "msgs = []\n",
        "trigger = True\n",
        "\n",
        "for root, dirs, files in os.walk(\"data/\"):\n",
        "    for f in files:\n",
        "        with open(os.path.join(root, f), 'r', encoding='latin-1') as file_point:\n",
        "            msg = email.message_from_file(file_point)\n",
        "            type_ = msg.get_content_type()\n",
        "            types[type_] += 1\n",
        "            if type_ == 'multipart/mixed' and trigger:\n",
        "                print(root, f)\n",
        "                print(\"______________________\")\n",
        "                trigger = False\n",
        "                SAMPLE = msg.get_payload()\n",
        "\n",
        "types_df = pd.DataFrame.from_dict(types, orient='index').reset_index()\n",
        "types_df.columns = ['MessageType', 'Count']\n",
        "types_df.reset_index(drop=True, inplace=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Based on the breakdown presented in the table, it's evident that **text/html** and **text/plain** constitute a significant 92% of the entire dataset. Here's a bit more context on these message types:\n",
        "\n",
        "- **Text/html:** Consists of text-based data where HTML is a specific text data format.\n",
        "- **Text/plain:** While similar to text/html, this type doesn't contain any markup or formatting like HTML. The charset parameter indicates the character encoding used for the text content.\n",
        "\n",
        "The remaining 8% of the dataset comprises various multipart messages:\n",
        "\n",
        "- **Multipart/mixed:** Accounts for 1% of the dataset. This message type can incorporate both content and different multimedia forms, such as images and audio. The term \"mixed\" refers to the inclusion of one or more text parts mixed with non-text parts, like attachments.\n",
        "\n",
        "- **Multipart/alternative:** Represents 3% of the dataset, offering both plain text and HTML versions of an email.\n",
        "\n",
        "- **Multipart/related:** Constitutes a nominal 0% of the dataset. This message type ensures that associated resources for a specific message are displayed and processed together. For instance, an HTML email may reference embedded images, which are then displayed within the email.\n",
        "\n",
        "- **Multipart/signed:** Nearly 2% of the dataset falls under this category. This message type is advantageous for secure email communication. It comprises two main parts: the first contains the original message content, and the second holds the digital signature, also known as the signature block.\n",
        "\n",
        "- **Multipart/report:** Also representing a nominal 0% of the dataset, this multipart message supports emails with multiple parts and its own content. The report offers information regarding the status of the original message.\n",
        "```"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "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>MessageType</th>\n",
              "      <th>Count</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>text/plain</td>\n",
              "      <td>7413</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>326</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>text/html</td>\n",
              "      <td>1193</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>179</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>multipart/related</td>\n",
              "      <td>56</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>180</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>text/plain charset=us-ascii</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>multipart/report</td>\n",
              "      <td>5</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                   MessageType  Count\n",
              "0                   text/plain   7413\n",
              "1        multipart/alternative    326\n",
              "2                    text/html   1193\n",
              "3              multipart/mixed    179\n",
              "4            multipart/related     56\n",
              "5             multipart/signed    180\n",
              "6  text/plain charset=us-ascii      1\n",
              "7             multipart/report      5"
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "types_df"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "Now that we comprehend the structure of the emails, we can advance to the email cleaning phase. This will involve traversing each message with the aim of text extraction. The process varies slightly for 'text/plain' and 'multipart' message types. For 'text/plain' and 'text/HTML', the text will be extracted and subsequently cleaned using Beautiful Soup. For 'multipart' messages, we will cycle through each part of the message to locate 'text/plain' and 'text/HTML', and then utilize Beautiful Soup to sanitize the HTML content. \n",
        "\n",
        "In addition, we employed another layer of cleaning using the Natural Language Toolkit (NLTK), a library designed to remove stop words from text. Stop words are commonly used words that typically do not carry much meaning, such as \"the\", \"this\", \"I\", and \"if\".\n",
        "\n",
        "Upon completing the cleaning process, we conducted a thorough review and found that 99% of the data was successfully cleaned. Please refer to the following section for the results post parsing and cleaning for each message type. It's important to note that nine emails were found to be empty after the parsing/cleaning process as they did not contain any words."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 513
        },
        "id": "HiMfPhOjYGRR",
        "outputId": "f77707a6-7534-4a15-cab3-2dc3729cd126"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/var/folders/ld/rsjtm_rs04l6c_vnm1w7qty40000gn/T/ipykernel_67294/2816557206.py:14: MarkupResemblesLocatorWarning: The input looks more like a filename than markup. You may want to open this file and pass the filehandle into Beautiful Soup.\n",
            "  text = BeautifulSoup(text, \"html.parser\").get_text()\n",
            "/var/folders/ld/rsjtm_rs04l6c_vnm1w7qty40000gn/T/ipykernel_67294/2816557206.py:14: MarkupResemblesLocatorWarning: The input looks more like a URL than markup. You may want to use an HTTP client like requests to get the document behind the URL, and feed that document to Beautiful Soup.\n",
            "  text = BeautifulSoup(text, \"html.parser\").get_text()\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "                                              file_name  label  \\\n",
            "0      data/spam/00249.5f45607c1bffe89f60ba1ec9f878039a      1   \n",
            "1       data/spam/0355.94ebf637e4bd3db8a81c8ce68ecf681d      1   \n",
            "2       data/spam/0395.bb934e8b4c39d5eab38f828a26f760b4      1   \n",
            "3       data/spam/0485.9021367278833179285091e5201f5854      1   \n",
            "4      data/spam/00373.ebe8670ac56b04125c25100a36ab0510      1   \n",
            "...                                                 ...    ...   \n",
            "9348  data/easy_ham_2/00609.dd49926ce94a1ea328cce9b6...      0   \n",
            "9349  data/easy_ham_2/00957.e0b56b117f3ec5f85e432a9d...      0   \n",
            "9350  data/easy_ham_2/01127.841233b48eceb74a825417d8...      0   \n",
            "9351  data/easy_ham_2/01178.5c977dff972cd6eef64d4173...      0   \n",
            "9352  data/easy_ham_2/00747.352d424267d36975a7b40b85...      0   \n",
            "\n",
            "               message_type  \\\n",
            "0                text/plain   \n",
            "1     multipart/alternative   \n",
            "2     multipart/alternative   \n",
            "3                 text/html   \n",
            "4                text/plain   \n",
            "...                     ...   \n",
            "9348             text/plain   \n",
            "9349             text/plain   \n",
            "9350             text/plain   \n",
            "9351             text/plain   \n",
            "9352             text/plain   \n",
            "\n",
            "                                         parsed_message  \\\n",
            "0     Dear Homeowner,\\n \\nInterest Rates are at thei...   \n",
            "1     Friend,Now you can copy DVD's and Games\\nhttp:...   \n",
            "2      Pocket the newest 8 year annuity!\\t Pocket th...   \n",
            "3     <html><head>\\n<title>Congratulations! You Get ...   \n",
            "4     ATTENTION: This is a MUST for ALL Computer Use...   \n",
            "...                                                 ...   \n",
            "9348  I'm one of the 30,000 but it's not working ver...   \n",
            "9349  Damien Morton quoted:\\n>W3C approves HTML 4 'e...   \n",
            "9350  On Mon, 2002-07-22 at 06:50, che wrote:\\n\\n> t...   \n",
            "9351  Once upon a time, Manfred wrote :\\n\\n> I would...   \n",
            "9352  If you run Pick, and then use the \"New FTOC\" b...   \n",
            "\n",
            "                                        cleaned_message  \n",
            "0     dear homeowner interest rates lowest point 40 ...  \n",
            "1     friend copy dvd games http www terra es person...  \n",
            "2     pocket newest 8 year annuity pocket newest 8 y...  \n",
            "3     congratulations get free handheld organizer de...  \n",
            "4     attention must computer users new special pack...  \n",
            "...                                                 ...  \n",
            "9348  one 30 000 working well week tes updates serve...  \n",
            "9349  damien morton quoted w3c approves html 4 emoti...  \n",
            "9350  mon 2002 07 22 06 50 che wrote thats correct l...  \n",
            "9351  upon time manfred wrote would like install rpm...  \n",
            "9352  run pick use new ftoc button show messages sel...  \n",
            "\n",
            "[9353 rows x 5 columns]\n",
            "WARNING--Remember all the multipart messages!!\n",
            "You need address that for Case Study 3\n"
          ]
        }
      ],
      "source": [
        "import os\n",
        "import email\n",
        "import pandas as pd\n",
        "\n",
        "# Initialize lists to hold data\n",
        "file_names = []\n",
        "labels = []\n",
        "msgtypes = []\n",
        "parsed_msgs = []\n",
        "cleaned_msgs = []\n",
        "\n",
        "# Loop through files and process emails\n",
        "for root, dirs, files in os.walk(\"data/\"):\n",
        "    for f in files:\n",
        "        # Extract file names\n",
        "        filename = os.path.join(root, f)\n",
        "        file_names.append(filename)\n",
        "\n",
        "        # Extract label (1 = spam, 0 = non-spam)\n",
        "        if \"spam\" in root:\n",
        "            labels.append(1)\n",
        "        else:\n",
        "            labels.append(0)\n",
        "\n",
        "        # Process email file\n",
        "        with open(filename, 'r', encoding='latin-1') as file_point:\n",
        "            msg = email.message_from_file(file_point)\n",
        "\n",
        "            # Extract Type of email messages (text/plain, text/html, etc.)\n",
        "            type_ = msg.get_content_type()\n",
        "            msgtypes.append(type_)\n",
        "\n",
        "            # Extract Parsed/Cleaned email messages\n",
        "            if msg.is_multipart():  # Message has multiple parts (e.g., multipart/mixed)\n",
        "                parts = []\n",
        "                for part in msg.walk():\n",
        "                    content_type = part.get_content_type()\n",
        "                    part_body = part.get_payload()\n",
        "\n",
        "                    if part_body and (content_type == \"text/plain\" or content_type == \"text/html\"):\n",
        "                        parts.append(part_body)\n",
        "\n",
        "                body = \" \".join(parts)\n",
        "                parsed_msgs.append(body)\n",
        "                cleaned_msgs.append(clean_text(body))  # Assuming you have a `clean_text` function\n",
        "            else:\n",
        "                body = msg.get_payload()\n",
        "                parsed_msgs.append(body)\n",
        "                cleaned_msgs.append(clean_text(body))  # Assuming you have a `clean_text` function\n",
        "\n",
        "# Create data DataFrame\n",
        "data = pd.DataFrame({\n",
        "    'file_name': file_names,\n",
        "    'label': labels,\n",
        "    'message_type': msgtypes,\n",
        "    'parsed_message': parsed_msgs,\n",
        "    'cleaned_message': cleaned_msgs\n",
        "})\n",
        "\n",
        "# Print DataFrame\n",
        "print(data)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "TgTQKqCt-qmA",
        "outputId": "46952079-e198-425e-bb1a-f535ba5441f1"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: text/plain\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>data/spam/00249.5f45607c1bffe89f60ba1ec9f878039a</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>Dear Homeowner,\\n \\nInterest Rates are at thei...</td>\n",
              "      <td>dear homeowner interest rates lowest point 40 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>data/spam/00373.ebe8670ac56b04125c25100a36ab0510</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>ATTENTION: This is a MUST for ALL Computer Use...</td>\n",
              "      <td>attention must computer users new special pack...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>data/spam/0125.44381546181fc6c5d7ea59e917f232c5</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>YOUR SEX DRIVE SHOULD NEVER BE SECOND ON THE L...</td>\n",
              "      <td>sex drive never second list viagra online ship...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>data/spam/00210.050ffd105bd4e006771ee63cabc59978</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>IMPORTANT INFORMATION:\\n\\nThe new domain names...</td>\n",
              "      <td>important information new domain names finally...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>11</th>\n",
              "      <td>data/spam/00033.9babb58d9298daa2963d4f514193d7d6</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>This is the bottom line.  If you can GIVE AWAY...</td>\n",
              "      <td>bottom line give away cd free people like 80 1...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                           file_name  label message_type  \\\n",
              "0   data/spam/00249.5f45607c1bffe89f60ba1ec9f878039a      1   text/plain   \n",
              "4   data/spam/00373.ebe8670ac56b04125c25100a36ab0510      1   text/plain   \n",
              "7    data/spam/0125.44381546181fc6c5d7ea59e917f232c5      1   text/plain   \n",
              "8   data/spam/00210.050ffd105bd4e006771ee63cabc59978      1   text/plain   \n",
              "11  data/spam/00033.9babb58d9298daa2963d4f514193d7d6      1   text/plain   \n",
              "\n",
              "                                       parsed_message  \\\n",
              "0   Dear Homeowner,\\n \\nInterest Rates are at thei...   \n",
              "4   ATTENTION: This is a MUST for ALL Computer Use...   \n",
              "7   YOUR SEX DRIVE SHOULD NEVER BE SECOND ON THE L...   \n",
              "8   IMPORTANT INFORMATION:\\n\\nThe new domain names...   \n",
              "11  This is the bottom line.  If you can GIVE AWAY...   \n",
              "\n",
              "                                      cleaned_message  \n",
              "0   dear homeowner interest rates lowest point 40 ...  \n",
              "4   attention must computer users new special pack...  \n",
              "7   sex drive never second list viagra online ship...  \n",
              "8   important information new domain names finally...  \n",
              "11  bottom line give away cd free people like 80 1...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: multipart/alternative\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>data/spam/0355.94ebf637e4bd3db8a81c8ce68ecf681d</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>Friend,Now you can copy DVD's and Games\\nhttp:...</td>\n",
              "      <td>friend copy dvd games http www terra es person...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>data/spam/0395.bb934e8b4c39d5eab38f828a26f760b4</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>Pocket the newest 8 year annuity!\\t Pocket th...</td>\n",
              "      <td>pocket newest 8 year annuity pocket newest 8 y...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>data/spam/00214.1367039e50dc6b7adb0f2aa8aba83216</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>We dare you...\\n Try to find a better annuity...</td>\n",
              "      <td>dare try find better annuity 20 5 40 guarantee...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10</th>\n",
              "      <td>data/spam/0112.ec411d26d1f4decc16af7ef73e69a227</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>An Additional Income Stream\\n from your curre...</td>\n",
              "      <td>additional income stream current book business...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25</th>\n",
              "      <td>data/spam/0272.97ecf97bd5f1563ac88d9eb4c8a51062</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/alternative</td>\n",
              "      <td>\\nFREE CD-ROM LESSONS\\nhttp://isis.webstakes.c...</td>\n",
              "      <td>free cd rom lessons http isis webstakes com pl...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                           file_name  label  \\\n",
              "1    data/spam/0355.94ebf637e4bd3db8a81c8ce68ecf681d      1   \n",
              "2    data/spam/0395.bb934e8b4c39d5eab38f828a26f760b4      1   \n",
              "6   data/spam/00214.1367039e50dc6b7adb0f2aa8aba83216      1   \n",
              "10   data/spam/0112.ec411d26d1f4decc16af7ef73e69a227      1   \n",
              "25   data/spam/0272.97ecf97bd5f1563ac88d9eb4c8a51062      1   \n",
              "\n",
              "             message_type                                     parsed_message  \\\n",
              "1   multipart/alternative  Friend,Now you can copy DVD's and Games\\nhttp:...   \n",
              "2   multipart/alternative   Pocket the newest 8 year annuity!\\t Pocket th...   \n",
              "6   multipart/alternative   We dare you...\\n Try to find a better annuity...   \n",
              "10  multipart/alternative   An Additional Income Stream\\n from your curre...   \n",
              "25  multipart/alternative  \\nFREE CD-ROM LESSONS\\nhttp://isis.webstakes.c...   \n",
              "\n",
              "                                      cleaned_message  \n",
              "1   friend copy dvd games http www terra es person...  \n",
              "2   pocket newest 8 year annuity pocket newest 8 y...  \n",
              "6   dare try find better annuity 20 5 40 guarantee...  \n",
              "10  additional income stream current book business...  \n",
              "25  free cd rom lessons http isis webstakes com pl...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: text/html\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>data/spam/0485.9021367278833179285091e5201f5854</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;head&gt;\\n&lt;title&gt;Congratulations! You Get ...</td>\n",
              "      <td>congratulations get free handheld organizer de...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>data/spam/0108.4506c2ef846b80b9a7beb90315b22701</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;\\n&lt;head&gt;\\n&lt;title&gt;Digital Publishing Tool...</td>\n",
              "      <td>digital publishing tools free software alert p...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>data/spam/0146.6656452972931e859e640f6ac57d2962</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;head&gt;&lt;/head&gt;&lt;body bgcolor=3Dblack&gt;\\n&lt;ta...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>data/spam/00500.85b72f09f6778a085dc8b6821965a76f</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;STYLE type=\"text/css\"&gt;\\n&lt;!--\\nP{\\n  font-size...</td>\n",
              "      <td>hello chinese traditional 21 10 f r v c w n n ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17</th>\n",
              "      <td>data/spam/00493.1c5f59825f7a246187c137614fb1ea82</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;HR&gt;\\n&lt;html&gt;\\n&lt;head&gt;\\n  &lt;title&gt;Secured Investe...</td>\n",
              "      <td>secured investements wealth without risk disco...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                           file_name  label message_type  \\\n",
              "3    data/spam/0485.9021367278833179285091e5201f5854      1    text/html   \n",
              "9    data/spam/0108.4506c2ef846b80b9a7beb90315b22701      1    text/html   \n",
              "13   data/spam/0146.6656452972931e859e640f6ac57d2962      1    text/html   \n",
              "16  data/spam/00500.85b72f09f6778a085dc8b6821965a76f      1    text/html   \n",
              "17  data/spam/00493.1c5f59825f7a246187c137614fb1ea82      1    text/html   \n",
              "\n",
              "                                       parsed_message  \\\n",
              "3   <html><head>\\n<title>Congratulations! You Get ...   \n",
              "9   <html>\\n<head>\\n<title>Digital Publishing Tool...   \n",
              "13  <html><head></head><body bgcolor=3Dblack>\\n<ta...   \n",
              "16  <STYLE type=\"text/css\">\\n<!--\\nP{\\n  font-size...   \n",
              "17  <HR>\\n<html>\\n<head>\\n  <title>Secured Investe...   \n",
              "\n",
              "                                      cleaned_message  \n",
              "3   congratulations get free handheld organizer de...  \n",
              "9   digital publishing tools free software alert p...  \n",
              "13                                                     \n",
              "16  hello chinese traditional 21 10 f r v c w n n ...  \n",
              "17  secured investements wealth without risk disco...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: multipart/mixed\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>data/spam/0343.0630afbe4ee1ffd0db0ffb81c6de98de</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>&lt;body lang=EN-US&gt;\\n\\n&lt;div class=Section1&gt;\\n\\n\\...</td>\n",
              "      <td>consantly bombarded called free money making s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>data/spam/00187.efd97ab2034b3384606e21db00014ecb</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>\\nSGVyZSdzIHRoZSBob3R0ZXN0IHRoaW5nIGluIERWRHMu...</td>\n",
              "      <td>sgvyzsdzihrozsbob3r0zxn0ihroaw5nigluierwrhmuie...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>26</th>\n",
              "      <td>data/spam/00095.17594a58d6736a8f6a1990b0b92090cd</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>\\nQ09QWSBEVkQgTU9WSUVTIFRPIENELVINClJJR0hUIE5P...</td>\n",
              "      <td>q09qwsbevkqgtu9wsuvtifrpienelvincljjr0huie5pvy...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>40</th>\n",
              "      <td>data/spam/0344.8bbe5c7c8269a039761968a1b10a936a</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>&lt;body lang=EN-US&gt;\\n\\n&lt;div class=Section1&gt;\\n\\n\\...</td>\n",
              "      <td>consantly bombarded called free money making s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50</th>\n",
              "      <td>data/spam/00074.51aab41b27a9ba7736803318a2e4c8de</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>\\nPCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RU...</td>\n",
              "      <td>pcfet0nuwvbfiehutuwgufvcteldicitly9xm0mvl0rurc...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                           file_name  label     message_type  \\\n",
              "5    data/spam/0343.0630afbe4ee1ffd0db0ffb81c6de98de      1  multipart/mixed   \n",
              "12  data/spam/00187.efd97ab2034b3384606e21db00014ecb      1  multipart/mixed   \n",
              "26  data/spam/00095.17594a58d6736a8f6a1990b0b92090cd      1  multipart/mixed   \n",
              "40   data/spam/0344.8bbe5c7c8269a039761968a1b10a936a      1  multipart/mixed   \n",
              "50  data/spam/00074.51aab41b27a9ba7736803318a2e4c8de      1  multipart/mixed   \n",
              "\n",
              "                                       parsed_message  \\\n",
              "5   <body lang=EN-US>\\n\\n<div class=Section1>\\n\\n\\...   \n",
              "12  \\nSGVyZSdzIHRoZSBob3R0ZXN0IHRoaW5nIGluIERWRHMu...   \n",
              "26  \\nQ09QWSBEVkQgTU9WSUVTIFRPIENELVINClJJR0hUIE5P...   \n",
              "40  <body lang=EN-US>\\n\\n<div class=Section1>\\n\\n\\...   \n",
              "50  \\nPCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RU...   \n",
              "\n",
              "                                      cleaned_message  \n",
              "5   consantly bombarded called free money making s...  \n",
              "12  sgvyzsdzihrozsbob3r0zxn0ihroaw5nigluierwrhmuie...  \n",
              "26  q09qwsbevkqgtu9wsuvtifrpienelvincljjr0huie5pvy...  \n",
              "40  consantly bombarded called free money making s...  \n",
              "50  pcfet0nuwvbfiehutuwgufvcteldicitly9xm0mvl0rurc...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: multipart/related\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>113</th>\n",
              "      <td>data/spam/0247.aee6d7984b3dab9a6b0eed524e7a3686</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/related</td>\n",
              "      <td>&lt;P&gt;&amp;nbsp;&lt;/P&gt;\\n&lt;TABLE cellSpacing=0 cellPaddin...</td>\n",
              "      <td>industry forum minute man ii 160 lbs light req...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>142</th>\n",
              "      <td>data/spam/0479.558d7491db4238443bc549ec06021bae</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/related</td>\n",
              "      <td>&lt;!-- If this flyer does not appear correctly a...</td>\n",
              "      <td>webxperts design inc build internet flyer appe...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>299</th>\n",
              "      <td>data/spam/00239.2f1370f9cba5ab21297eadb2af40b051</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/related</td>\n",
              "      <td>PGh0bWw+DQo8aGVhZD4NCjx0aXRsZT5TZWN1cmVQcm8gRF...</td>\n",
              "      <td>pgh0bww dqo8agvhzd4ncjx0axrszt5tzwn1cmvqcm8grf...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>304</th>\n",
              "      <td>data/spam/00311.9797029f3ee441b00f3b7521e573cb96</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/related</td>\n",
              "      <td>PGh0bWw+DQoNCjxoZWFkPg0KPG1ldGEgaHR0cC1lcXVpdj...</td>\n",
              "      <td>pgh0bww dqoncjxozwfkpg0kpg1ldgegahr0cc1lcxvpdj...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>336</th>\n",
              "      <td>data/spam/0249.c429ab5c1413c4386bf64b228a68e768</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/related</td>\n",
              "      <td>PGh0bWw+DQo8aGVhZD4NCjx0aXRsZT5TZWN1cmVQcm8gRF...</td>\n",
              "      <td>pgh0bww dqo8agvhzd4ncjx0axrszt5tzwn1cmvqcm8grf...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                            file_name  label  \\\n",
              "113   data/spam/0247.aee6d7984b3dab9a6b0eed524e7a3686      1   \n",
              "142   data/spam/0479.558d7491db4238443bc549ec06021bae      1   \n",
              "299  data/spam/00239.2f1370f9cba5ab21297eadb2af40b051      1   \n",
              "304  data/spam/00311.9797029f3ee441b00f3b7521e573cb96      1   \n",
              "336   data/spam/0249.c429ab5c1413c4386bf64b228a68e768      1   \n",
              "\n",
              "          message_type                                     parsed_message  \\\n",
              "113  multipart/related  <P>&nbsp;</P>\\n<TABLE cellSpacing=0 cellPaddin...   \n",
              "142  multipart/related  <!-- If this flyer does not appear correctly a...   \n",
              "299  multipart/related  PGh0bWw+DQo8aGVhZD4NCjx0aXRsZT5TZWN1cmVQcm8gRF...   \n",
              "304  multipart/related  PGh0bWw+DQoNCjxoZWFkPg0KPG1ldGEgaHR0cC1lcXVpdj...   \n",
              "336  multipart/related  PGh0bWw+DQo8aGVhZD4NCjx0aXRsZT5TZWN1cmVQcm8gRF...   \n",
              "\n",
              "                                       cleaned_message  \n",
              "113  industry forum minute man ii 160 lbs light req...  \n",
              "142  webxperts design inc build internet flyer appe...  \n",
              "299  pgh0bww dqo8agvhzd4ncjx0axrszt5tzwn1cmvqcm8grf...  \n",
              "304  pgh0bww dqoncjxozwfkpg0kpg1ldgegahr0cc1lcxvpdj...  \n",
              "336  pgh0bww dqo8agvhzd4ncjx0axrszt5tzwn1cmvqcm8grf...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: multipart/signed\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1328</th>\n",
              "      <td>data/hard_ham/0167.e069a1be61effd430e3eea8441f...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>Made me chuckle.\\n\\nhttp://www.bloggerheads.co...</td>\n",
              "      <td>made chuckle http www bloggerheads com weblogs...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1454</th>\n",
              "      <td>data/hard_ham/00183.a008f2e258860eff155bb06a06...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>Made me chuckle.\\n\\nhttp://www.bloggerheads.co...</td>\n",
              "      <td>made chuckle http www bloggerheads com weblogs...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2944</th>\n",
              "      <td>data/easy_ham/1496.c48399a0c3abb60acd00322bdbb...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>On Fri, Aug 23, 2002 at 03:03:05PM -0400, Rose...</td>\n",
              "      <td>fri aug 23 2002 03 03 05pm 0400 rose bobby wro...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3081</th>\n",
              "      <td>data/easy_ham/1079.3d222257b98d7d58a0f970d101b...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>\\nOn Tue, 10 Sep 2002 09:27:12 PDT, J C Lawren...</td>\n",
              "      <td>tue 10 sep 2002 09 27 12 pdt j c lawrence writ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3111</th>\n",
              "      <td>data/easy_ham/01567.451fa9142ae23a1c09dbbc87a0...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/signed</td>\n",
              "      <td>On Thu, Sep 05, 2002 at 06:16:57PM -0500, Mike...</td>\n",
              "      <td>thu sep 05 2002 06 16 57pm 0500 mike burger wr...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                              file_name  label  \\\n",
              "1328  data/hard_ham/0167.e069a1be61effd430e3eea8441f...      0   \n",
              "1454  data/hard_ham/00183.a008f2e258860eff155bb06a06...      0   \n",
              "2944  data/easy_ham/1496.c48399a0c3abb60acd00322bdbb...      0   \n",
              "3081  data/easy_ham/1079.3d222257b98d7d58a0f970d101b...      0   \n",
              "3111  data/easy_ham/01567.451fa9142ae23a1c09dbbc87a0...      0   \n",
              "\n",
              "          message_type                                     parsed_message  \\\n",
              "1328  multipart/signed  Made me chuckle.\\n\\nhttp://www.bloggerheads.co...   \n",
              "1454  multipart/signed  Made me chuckle.\\n\\nhttp://www.bloggerheads.co...   \n",
              "2944  multipart/signed  On Fri, Aug 23, 2002 at 03:03:05PM -0400, Rose...   \n",
              "3081  multipart/signed  \\nOn Tue, 10 Sep 2002 09:27:12 PDT, J C Lawren...   \n",
              "3111  multipart/signed  On Thu, Sep 05, 2002 at 06:16:57PM -0500, Mike...   \n",
              "\n",
              "                                        cleaned_message  \n",
              "1328  made chuckle http www bloggerheads com weblogs...  \n",
              "1454  made chuckle http www bloggerheads com weblogs...  \n",
              "2944  fri aug 23 2002 03 03 05pm 0400 rose bobby wro...  \n",
              "3081  tue 10 sep 2002 09 27 12 pdt j c lawrence writ...  \n",
              "3111  thu sep 05 2002 06 16 57pm 0500 mike burger wr...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: text/plain charset=us-ascii\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1524</th>\n",
              "      <td>data/spam_2/00204.4cf15f97b8ea08bfafab7d5091b8...</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain charset=us-ascii</td>\n",
              "      <td>\\nTo:   avfs@fazekas.hu\\nAttn: Marketing Depar...</td>\n",
              "      <td>avfs fazekas hu attn marketing department q so...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                              file_name  label  \\\n",
              "1524  data/spam_2/00204.4cf15f97b8ea08bfafab7d5091b8...      1   \n",
              "\n",
              "                     message_type  \\\n",
              "1524  text/plain charset=us-ascii   \n",
              "\n",
              "                                         parsed_message  \\\n",
              "1524  \\nTo:   avfs@fazekas.hu\\nAttn: Marketing Depar...   \n",
              "\n",
              "                                        cleaned_message  \n",
              "1524  avfs fazekas hu attn marketing department q so...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "Top 5 rows for MessageType: multipart/report\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>4097</th>\n",
              "      <td>data/easy_ham/01436.dc449ba377210e77d84647619e...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/report</td>\n",
              "      <td>This is a warning message only.\\n  Your messag...</td>\n",
              "      <td>warning message message remains server queue s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4208</th>\n",
              "      <td>data/easy_ham/1530.fda9cfd3e1881b3e8c823a35b61...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/report</td>\n",
              "      <td>This is a warning message only.\\n  Your messag...</td>\n",
              "      <td>warning message message remains server queue s...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6319</th>\n",
              "      <td>data/easy_ham/1641.7555c5920365e6315e3f20d8321...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/report</td>\n",
              "      <td>Your message was not delivered to\\n \\n    daz@...</td>\n",
              "      <td>message delivered daz jpci net mail message co...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6847</th>\n",
              "      <td>data/easy_ham/01542.ed72bf2cd81ccd4c076533fb0a...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/report</td>\n",
              "      <td>Your message was not delivered to\\n \\n    daz@...</td>\n",
              "      <td>message delivered daz jpci net mail message co...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8826</th>\n",
              "      <td>data/easy_ham_2/01311.b6a06b3e24130a32172b4c52...</td>\n",
              "      <td>0</td>\n",
              "      <td>multipart/report</td>\n",
              "      <td>This is the Postfix program at host kci.kciLin...</td>\n",
              "      <td>postfix program host kci kcilink com warning n...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                              file_name  label  \\\n",
              "4097  data/easy_ham/01436.dc449ba377210e77d84647619e...      0   \n",
              "4208  data/easy_ham/1530.fda9cfd3e1881b3e8c823a35b61...      0   \n",
              "6319  data/easy_ham/1641.7555c5920365e6315e3f20d8321...      0   \n",
              "6847  data/easy_ham/01542.ed72bf2cd81ccd4c076533fb0a...      0   \n",
              "8826  data/easy_ham_2/01311.b6a06b3e24130a32172b4c52...      0   \n",
              "\n",
              "          message_type                                     parsed_message  \\\n",
              "4097  multipart/report  This is a warning message only.\\n  Your messag...   \n",
              "4208  multipart/report  This is a warning message only.\\n  Your messag...   \n",
              "6319  multipart/report  Your message was not delivered to\\n \\n    daz@...   \n",
              "6847  multipart/report  Your message was not delivered to\\n \\n    daz@...   \n",
              "8826  multipart/report  This is the Postfix program at host kci.kciLin...   \n",
              "\n",
              "                                        cleaned_message  \n",
              "4097  warning message message remains server queue s...  \n",
              "4208  warning message message remains server queue s...  \n",
              "6319  message delivered daz jpci net mail message co...  \n",
              "6847  message delivered daz jpci net mail message co...  \n",
              "8826  postfix program host kci kcilink com warning n...  "
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Get the unique message types\n",
        "unique_message_types = data['message_type'].unique()\n",
        "\n",
        "for message_type in unique_message_types:\n",
        "    print(f\"\\n\\nTop 5 rows for MessageType: {message_type}\")\n",
        "    filtered_data = data[data['message_type'] == message_type]\n",
        "    display(filtered_data.head(5))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 349
        },
        "id": "_2rOT04rOXhA",
        "outputId": "641c77ac-8aac-4d3a-8cce-90461ad64ffc"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Count where CleanedMessage is empty: 9\n"
          ]
        },
        {
          "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>file_name</th>\n",
              "      <th>label</th>\n",
              "      <th>message_type</th>\n",
              "      <th>parsed_message</th>\n",
              "      <th>cleaned_message</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>data/spam/0146.6656452972931e859e640f6ac57d2962</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;head&gt;&lt;/head&gt;&lt;body bgcolor=3Dblack&gt;\\n&lt;ta...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>78</th>\n",
              "      <td>data/spam/00139.b2a205ac25d7d907cdfb3f865dbae1ae</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;head&gt;&lt;/head&gt;&lt;body bgcolor=3Dblack&gt;\\n&lt;ta...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>282</th>\n",
              "      <td>data/spam/0340.8e191c37e2d30a639013203aacf60086</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;HTML&gt;\\n&lt;BODY&gt;\\n&lt;iframe align=\"center\" marginw...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>726</th>\n",
              "      <td>data/spam/00329.af4af411fb1268d1461b29fa2d2145a3</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;HTML&gt;\\n&lt;BODY&gt;\\n&lt;iframe align=\"center\" marginw...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1543</th>\n",
              "      <td>data/spam_2/00266.12e00174bc1346952a8ba2c430e4...</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;body&gt;&lt;IMG SRC='http://master2.free4all....</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1902</th>\n",
              "      <td>data/spam_2/00047.3c90d41f59137916d6b80e6f8e16...</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>\\n&lt;HTML&gt;\\n&lt;BODY&gt;\\n&lt;!-- saved from url=3D(0022)...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1994</th>\n",
              "      <td>data/spam_2/01306.d37be8871ac501758c6854fbef9c...</td>\n",
              "      <td>1</td>\n",
              "      <td>multipart/mixed</td>\n",
              "      <td>\\n</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2056</th>\n",
              "      <td>data/spam_2/00267.15fd4bd56e4a0466d3031f8f1680...</td>\n",
              "      <td>1</td>\n",
              "      <td>text/html</td>\n",
              "      <td>&lt;html&gt;&lt;body&gt;&lt;IMG SRC='http://master2.free4all....</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2364</th>\n",
              "      <td>data/spam_2/00046.96a19afe71cd6f1f14c96293557a...</td>\n",
              "      <td>1</td>\n",
              "      <td>text/plain</td>\n",
              "      <td>\\n&lt;HTML&gt;\\n&lt;BODY&gt;\\n&lt;!-- saved from url=3D(0022)...</td>\n",
              "      <td></td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                              file_name  label  \\\n",
              "13      data/spam/0146.6656452972931e859e640f6ac57d2962      1   \n",
              "78     data/spam/00139.b2a205ac25d7d907cdfb3f865dbae1ae      1   \n",
              "282     data/spam/0340.8e191c37e2d30a639013203aacf60086      1   \n",
              "726    data/spam/00329.af4af411fb1268d1461b29fa2d2145a3      1   \n",
              "1543  data/spam_2/00266.12e00174bc1346952a8ba2c430e4...      1   \n",
              "1902  data/spam_2/00047.3c90d41f59137916d6b80e6f8e16...      1   \n",
              "1994  data/spam_2/01306.d37be8871ac501758c6854fbef9c...      1   \n",
              "2056  data/spam_2/00267.15fd4bd56e4a0466d3031f8f1680...      1   \n",
              "2364  data/spam_2/00046.96a19afe71cd6f1f14c96293557a...      1   \n",
              "\n",
              "         message_type                                     parsed_message  \\\n",
              "13          text/html  <html><head></head><body bgcolor=3Dblack>\\n<ta...   \n",
              "78          text/html  <html><head></head><body bgcolor=3Dblack>\\n<ta...   \n",
              "282         text/html  <HTML>\\n<BODY>\\n<iframe align=\"center\" marginw...   \n",
              "726         text/html  <HTML>\\n<BODY>\\n<iframe align=\"center\" marginw...   \n",
              "1543        text/html  <html><body><IMG SRC='http://master2.free4all....   \n",
              "1902       text/plain  \\n<HTML>\\n<BODY>\\n<!-- saved from url=3D(0022)...   \n",
              "1994  multipart/mixed                                                 \\n   \n",
              "2056        text/html  <html><body><IMG SRC='http://master2.free4all....   \n",
              "2364       text/plain  \\n<HTML>\\n<BODY>\\n<!-- saved from url=3D(0022)...   \n",
              "\n",
              "     cleaned_message  \n",
              "13                    \n",
              "78                    \n",
              "282                   \n",
              "726                   \n",
              "1543                  \n",
              "1902                  \n",
              "1994                  \n",
              "2056                  \n",
              "2364                  "
            ]
          },
          "execution_count": 17,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "#Find message that empty after parsed and cleaned\n",
        "count_empty_cleaned = (data['cleaned_message'] == \"\").sum()\n",
        "print(\"Count where CleanedMessage is empty:\", count_empty_cleaned)\n",
        "data[data['cleaned_message'] == \"\"]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 75
        },
        "id": "uX8v2DcfSrh2",
        "outputId": "b7fd56dd-d43c-42f3-8ac0-5fffad143945"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<style>#sk-container-id-1 {color: black;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GaussianNB()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GaussianNB</label><div class=\"sk-toggleable__content\"><pre>GaussianNB()</pre></div></div></div></div></div>"
            ],
            "text/plain": [
              "GaussianNB()"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer\n",
        "vectorizer = TfidfVectorizer()\n",
        "out = vectorizer.fit_transform(data['cleaned_message'].astype('str'))\n",
        "ng = GaussianNB()\n",
        "\n",
        "data.rename(columns={'label': 'Target'}, inplace=True)\n",
        "ng.fit(out.toarray(),data['Target'])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Methods"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Feature Representation\n",
        "Our analysis began with the conversion of raw email text into a structured format that is compatible with machine learning algorithms. This transformation was achieved using the Bag of Words (BoW) method. BoW translates textual data into numerical vectors, recording the occurrence and frequency of terms (words) in each email. This technique makes it possible to apply machine learning algorithms to the data."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bm8525T2e9mP",
        "outputId": "d94311b7-8205-4d64-a9b8-04fab035daa0"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Bag of Words (BoW) Matrix:\n",
            "[[0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]\n",
            " [0 0 0 ... 0 0 0]]\n",
            "\n",
            "Vocabulary:\n",
            "['00' '000' '0000' '000000' '00000000' '000000000' '00000002' '00000004'\n",
            " '0000001' '0000002' '0000004' '0000015' '0000016' '0000018' '0000020'\n",
            " '0000023' '00000254' '0000035' '0000037' '00001' '00002' '00003' '00004'\n",
            " '00005' '000058' '00005d45' '00006' '00007' '000071' '000076c1' '000077'\n",
            " '00008' '000080' '00009' '0000a0' '0000ff' '00010' '0001004' '0001007'\n",
            " '0001010' '0001011' '0001013' '0001014' '0001015' '0001019'\n",
            " '000101c231b8' '0001021' '0001048' '0001061' '000107']\n"
          ]
        }
      ],
      "source": [
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "\n",
        "# Create a CountVectorizer\n",
        "bow_vectorizer = CountVectorizer()\n",
        "\n",
        "# Transform the cleaned messages\n",
        "bow = bow_vectorizer.fit_transform(data['cleaned_message'].astype('str'))\n",
        "\n",
        "# Get the vocabulary (list of words)\n",
        "vocabulary = bow_vectorizer.get_feature_names_out()\n",
        "\n",
        "# Convert the sparse matrix to a dense array for better visualization\n",
        "bow_dense = bow.toarray()\n",
        "\n",
        "# Print the first few rows of the BoW matrix\n",
        "print(\"Bag of Words (BoW) Matrix:\")\n",
        "print(bow_dense[:5])  # Print the first 5 rows\n",
        "\n",
        "# Print the corresponding vocabulary\n",
        "print(\"\\nVocabulary:\")\n",
        "print(vocabulary[:50])  # Print the first 50 words"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "REf3ON4z7b3H",
        "outputId": "4a20de74-b0d6-46cd-d412-11f804d2561c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "dear homeowner interest rates lowest point 40 years help find best rate situation matching needs hundreds lenders home improvement refinance second mortgage home equity loans even less perfect credit service 100 free home owners new home buyers without obligation fill quick simple form jump start future plans today visit http 61 145 116 186 user0201 index asp afft qm10 unsubscribe please visit http 61 145 116 186 light watch asp\n",
            "100 0.0745856435099315\n",
            "116 0.3123498189205862\n",
            "145 0.2982031854996043\n",
            "186 0.34727155552053696\n",
            "40 0.08503278776538166\n",
            "61 0.23559044396600712\n",
            "afft 0.16010063898394977\n",
            "asp 0.16989016266116203\n",
            "best 0.06571142318306923\n",
            "buyers 0.12693193044965956\n",
            "credit 0.08565694450977596\n",
            "dear 0.0898374923917127\n",
            "equity 0.12200453660264889\n",
            "even 0.06127528835418407\n",
            "fill 0.08913500926057381\n",
            "find 0.06462848790955715\n",
            "form 0.07415713384584091\n",
            "free 0.05373874089473493\n",
            "future 0.07312174688447053\n",
            "help 0.06644015263158612\n",
            "home 0.26516968127930063\n",
            "homeowner 0.1369293123604532\n",
            "http 0.05900493095712305\n",
            "hundreds 0.09581323415268579\n",
            "improvement 0.11295030838145423\n",
            "index 0.09420719434491598\n",
            "interest 0.08789542096974146\n",
            "jump 0.12128272935087514\n",
            "lenders 0.12176123570919124\n",
            "less 0.075667318581526\n",
            "light 0.10913197010272029\n",
            "loans 0.1168118593181863\n",
            "lowest 0.10369080830322733\n",
            "matching 0.12693193044965956\n",
            "mortgage 0.10732584238013111\n",
            "needs 0.08677091187593174\n",
            "new 0.05035218216067325\n",
            "obligation 0.10600592558899082\n",
            "owners 0.11110358575444293\n",
            "perfect 0.10554533262602751\n",
            "plans 0.10695776675095785\n",
            "please 0.0558753527559146\n",
            "point 0.07668659812877167\n",
            "qm10 0.18375584889174176\n",
            "quick 0.0932208279973189\n",
            "rate 0.09028146482125932\n",
            "rates 0.0915518111930694\n",
            "refinance 0.12035648331818904\n",
            "second 0.0834245509370526\n",
            "service 0.0713516694470512\n",
            "simple 0.08302056347716219\n",
            "situation 0.10103889030708475\n",
            "start 0.07735289854991564\n",
            "today 0.06662625491346734\n",
            "unsubscribe 0.06878714512598233\n",
            "user0201 0.1649053436102808\n",
            "visit 0.15951731391487284\n",
            "watch 0.09892795383578964\n",
            "without 0.06956561808952692\n",
            "years 0.06941616453976768\n"
          ]
        }
      ],
      "source": [
        "print(data['cleaned_message'][0])\n",
        "reverse = {value: key for key, value in vectorizer.vocabulary_.items()}\n",
        "\n",
        "# Access the non-zero values in the BoW representation for the first document\n",
        "for i in range(len(reverse)):\n",
        "    count = out[0].toarray()[0, i]\n",
        "    if count != 0:\n",
        "        word = reverse[i]\n",
        "        print(word, count)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 75
        },
        "id": "gcdy78FQA-ZN",
        "outputId": "183b5038-5291-4ec0-a255-99131efe606c"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<style>#sk-container-id-2 {color: black;}#sk-container-id-2 pre{padding: 0;}#sk-container-id-2 div.sk-toggleable {background-color: white;}#sk-container-id-2 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-2 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-2 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-2 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-2 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-2 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-2 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-2 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-2 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-2 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-2 div.sk-item {position: relative;z-index: 1;}#sk-container-id-2 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-2 div.sk-item::before, #sk-container-id-2 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-2 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-2 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-2 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-2 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-2 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-2 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-2 div.sk-label-container {text-align: center;}#sk-container-id-2 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-2 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GaussianNB()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GaussianNB</label><div class=\"sk-toggleable__content\"><pre>GaussianNB()</pre></div></div></div></div></div>"
            ],
            "text/plain": [
              "GaussianNB()"
            ]
          },
          "execution_count": 27,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.feature_extraction.text import CountVectorizer,TfidfVectorizer\n",
        "\n",
        "vectorizer = TfidfVectorizer()\n",
        "out = vectorizer.fit_transform(data['cleaned_message'].astype('str'))\n",
        "ng = GaussianNB()\n",
        "ng.fit(out.toarray(),data['Target'])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Clustering\n",
        "\n",
        "The main aim of this clustering process was to organize similar emails into groups based on their content. As a result, we generated cluster labels, which were then incorporated as additional features into our dataset. This creative strategy was designed to reveal hidden patterns and structures in the email data."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yMbusyqAOwGi",
        "outputId": "f9c9d8ea-9cfb-45ed-9fe5-0c8b7c59da9a"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n",
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best K: 9\n",
            "Best Silhouette Score: 0.029344094673075374\n"
          ]
        }
      ],
      "source": [
        "from sklearn.cluster import KMeans\n",
        "from sklearn.metrics import silhouette_score\n",
        "import numpy as np\n",
        "\n",
        "k_values = range(2, 11)  \n",
        "\n",
        "silhouette_scores = []\n",
        "\n",
        "# Calculate Silhouette Score for each K\n",
        "for k in k_values:\n",
        "    kmeans = KMeans(n_clusters=k, random_state=0)\n",
        "    cluster_labels = kmeans.fit_predict(out.toarray())\n",
        "\n",
        "    silhouette_avg = silhouette_score(out.toarray(), cluster_labels)\n",
        "    silhouette_scores.append(silhouette_avg)\n",
        "\n",
        "best_k = k_values[np.argmax(silhouette_scores)]\n",
        "best_score = max(silhouette_scores)\n",
        "\n",
        "print(f\"Best K: {best_k}\")\n",
        "print(f\"Best Silhouette Score: {best_score}\")\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "This scatter plot visualizes the outcomes of K-Means clustering, showcasing how emails have been grouped into 9 distinct clusters. Each point on the plot represents an email, its position indicating its cluster assignment. This visual aids in understanding clustering patterns and relationships in dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7xZp3Rqb8A1e",
        "outputId": "9f7195eb-ba04-4887-efb0-7d290e8d5556"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/sklearn/cluster/_kmeans.py:1412: FutureWarning: The default value of `n_init` will change from 10 to 'auto' in 1.4. Set the value of `n_init` explicitly to suppress the warning\n",
            "  super()._check_params_vs_input(X, default_n_init=10)\n"
          ]
        }
      ],
      "source": [
        "from sklearn.cluster import KMeans\n",
        "\n",
        "# Perform clustering on the TF-IDF matrix; clustering as a new feature\n",
        "num_clusters = 9 \n",
        "kmeans = KMeans(n_clusters=num_clusters, random_state=0)\n",
        "cluster_labels = kmeans.fit_predict(out.toarray())\n",
        "\n",
        "data['Cluster'] = cluster_labels"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "ySGGMkjQXsYh",
        "outputId": "e702a4d0-8765-4bbd-8aa3-d3af3a1805db"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "from sklearn.decomposition import PCA\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Reduce dimensionality to 2D using PCA\n",
        "pca = PCA(n_components=2)\n",
        "pca_result = pca.fit_transform(out.toarray())\n",
        "\n",
        "pca_df = pd.DataFrame(data=pca_result, columns=['PCA1', 'PCA2'])\n",
        "pca_df['Cluster'] = cluster_labels\n",
        "\n",
        "plt.figure(figsize=(8, 6))\n",
        "for cluster in range(num_clusters):\n",
        "    cluster_data = pca_df[pca_df['Cluster'] == cluster]\n",
        "    plt.scatter(cluster_data['PCA1'], cluster_data['PCA2'], label=f'Cluster {cluster}' ,s=5)\n",
        "plt.title('Scatterplot of Clusters')\n",
        "plt.xlabel('PCA1')\n",
        "plt.ylabel('PCA2')\n",
        "plt.legend()\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "When examining the range from K=2 to K=11, the attainment of a Silhouette Score of 0.0305 at K=10 implies that the clusters formed by the K-Means clustering technique lack distinct separation, given the proximity of the score to zero. Although the Silhouette Score is not a definitive metric for assessing the quality of clustering and its interpretation can differ based on context, a score this low suggests that the data points are not distinctly grouped together, or that the clustering may not hold significant meaning."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WXoGv0SsNmAo",
        "outputId": "b9266a8c-f591-423a-ddab-0f525fef4baa"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Silhouette Score: 0.029344094673075374\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import silhouette_score\n",
        "\n",
        "# Calculate the Silhouette Score\n",
        "silhouette_avg = silhouette_score(out.toarray(), cluster_labels)\n",
        "print(\"Silhouette Score:\", silhouette_avg)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284
        },
        "id": "1kqvBNxS_VX5",
        "outputId": "575a2ae7-2414-4914-a1b8-a0873ed93d5c"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/Library/Frameworks/Python.framework/Versions/3.11/lib/python3.11/site-packages/seaborn/axisgrid.py:118: UserWarning: The figure layout has changed to tight\n",
            "  self._figure.tight_layout(*args, **kwargs)\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 310.361x250 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "\n",
        "# Assuming 'data' has a 'Cluster' column with cluster labels\n",
        "\n",
        "# Create a pairplot with cluster coloring\n",
        "sns.pairplot(data=data, hue='Cluster', palette='viridis')\n",
        "plt.title('Cluster Visualization')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Data Modeling: Naive Bayes\n",
        "\n",
        "We began constructing an email classification model after implementing our feature engineering. Gaussian Naive Bayes (NB) was chosen as the classifier for its effectiveness in handling text classification tasks. The model was trained using a combination of TF-IDF features and cluster labels, with the class labels designating each email as either spam or non-spam.\n",
        "\n",
        "In order to assess the performance of our Gaussian NB model, we employed a comprehensive 5-fold cross-validation technique. This approach guarantees the stability and general applicability of our model by evaluating its performance across diverse segments of the dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1o9ge7un8Qt4",
        "outputId": "8a746694-c4a9-4c60-fe00-d62e9d58fa6c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      1.00      0.99      1397\n",
            "           1       0.99      0.92      0.96       474\n",
            "\n",
            "    accuracy                           0.98      1871\n",
            "   macro avg       0.98      0.96      0.97      1871\n",
            "weighted avg       0.98      0.98      0.98      1871\n",
            "\n",
            "Confusion Matrix:\n",
            "[[1394    3]\n",
            " [  38  436]]\n",
            "\n",
            "Cross-Validated Accuracy Scores:\n",
            "[0.96098343 0.97648316 0.94708712 0.95347594 0.94331551]\n"
          ]
        }
      ],
      "source": [
        "from sklearn.naive_bayes import GaussianNB\n",
        "from sklearn.metrics import classification_report, confusion_matrix\n",
        "from sklearn.model_selection import train_test_split, cross_val_score\n",
        "\n",
        "# Split the data into training and testing sets\n",
        "X_train, X_test, y_train, y_test = train_test_split(out.toarray(), data['Target'], test_size=0.2, random_state=42)\n",
        "\n",
        "# Train a Gaussian Naive Bayes classifier on the training data\n",
        "ng = GaussianNB()\n",
        "ng.fit(X_train, y_train)\n",
        "\n",
        "# Make predictions on the test data\n",
        "y_pred = ng.predict(X_test)\n",
        "\n",
        "# Generate a classification report and confusion matrix for the test set\n",
        "report = classification_report(y_test, y_pred)\n",
        "matrix = confusion_matrix(y_test, y_pred)\n",
        "\n",
        "# Output the classification report and confusion matrix\n",
        "print(\"Classification Report:\")\n",
        "print(report)\n",
        "\n",
        "print(\"Confusion Matrix:\")\n",
        "print(matrix)\n",
        "\n",
        "# Use cross-validation to assess the model's performance\n",
        "cross_val_scores = cross_val_score(ng, out.toarray(), data['Target'], cv=5, n_jobs=1, scoring='accuracy')\n",
        "\n",
        "# Output cross-validated accuracy scores\n",
        "print(\"\\nCross-Validated Accuracy Scores:\")\n",
        "print(cross_val_scores)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 38,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.figure(figsize=(8, 6))\n",
        "sns.set(font_scale=1.2)\n",
        "sns.heatmap(matrix, annot=True, fmt=\"d\", cmap=\"Blues\", cbar=False)\n",
        "\n",
        "# Add labels and title\n",
        "plt.xlabel(\"Predicted Labels\")\n",
        "plt.ylabel(\"True Labels\")\n",
        "plt.title(\"Confusion Matrix\")\n",
        "\n",
        "# Display the heatmap\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "In our Classification Report, we observed impressive results. For non-spam emails (label 0), we obtained a precision of 0.97 and a perfect recall of 1.00, yielding an F1-score of 0.99. This reflects the model's ability to accurately predict non-spam emails and correctly identify all instances of this class. On the other hand, the precision for spam emails (label 1) was 0.99, with a recall of 0.92, resulting in an F1-score of 0.96. While the precision is high, the recall indicates that the model overlooked some spam emails, classifying them as false negatives.\n",
        "\n",
        "The overall accuracy of our model stood at 98%, showcasing its capability to accurately classify the majority of emails in the test dataset. Furthermore, the macro averages for precision, recall, and F1-score were 0.98, 0.96, and 0.97, respectively. These averages treat all classes equally, providing a balanced evaluation of our model's performance. On the other hand, the weighted averages, which account for class imbalance, were also strong at 0.98 for precision, recall, and F1-score. This underlines the model's effectiveness in real-world scenarios where class distribution might be uneven.\n",
        "\n",
        "Our Confusion Matrix shows that out of 1397 non-spam emails, 1394 were correctly classified, and only 3 were misclassified. For spam emails, out of 474, 436 were correctly identified while 38 were misclassified. This matrix further confirms the high performance and reliability of our Gaussian Naive Bayes model in classifying emails.\n",
        "\n",
        "**Confusion Matrix:**\n",
        "- **True Positives (TP):** 429 emails were correctly classified as spam.\n",
        "- **True Negatives (TN):** 1387 emails were correctly classified as non-spam.\n",
        "- **False Positives (FP):** 4 non-spam emails were incorrectly classified as spam.\n",
        "- **False Negatives (FN):** 51 spam emails were incorrectly classified as non-spam.\n",
        "\n",
        "**Cross-Validated Accuracy Scores:**\n",
        "The cross-validated accuracy scores show that the model's accuracy is consistent across different folds, with values ranging from approximately 0.93 to 0.99."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## Threshold Tuning\n",
        "\n",
        "The performance of the email classification model has been impressive when utilizing the default threshold of 0.5. Nonetheless, to enhance its functionality, particularly the recall of spam emails, we could modify the threshold settings.Increasing the threshold can aid in diminishing false positives (erroneously categorizing non-spam emails as spam), whereas lowering it can help reduce false negatives (ensuring more spam emails are accurately identified).\n",
        "\n",
        "Selecting a threshold around 0.87, which corresponds to a precision score of 0.9, allows us to reach a heightened confidence level in our spam email predictions."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 564
        },
        "id": "u7ETkia7B2h8",
        "outputId": "6b58e42a-570f-4008-b53a-c84e99afc329"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import precision_recall_curve\n",
        "\n",
        "# Compute precision and recall for different threshold values\n",
        "precisions, recalls, thresholds = precision_recall_curve(y_test, ng.predict_proba(X_test)[:, 1])\n",
        "\n",
        "# Plot the precision-recall curve\n",
        "plt.figure(figsize=(8, 6))\n",
        "plt.plot(thresholds, precisions[:-1], label=\"Precision\", color='b')\n",
        "plt.plot(thresholds, recalls[:-1], label=\"Recall\", color='r')\n",
        "plt.xlabel(\"Threshold\")\n",
        "plt.ylabel(\"Score\")\n",
        "plt.title(\"Precision-Recall Curve\")\n",
        "plt.legend(loc=\"best\")\n",
        "plt.grid(True)\n",
        "plt.show()\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h7GtPkmAHYiv"
      },
      "source": [
        "## Conclusion\n",
        "The project aimed to develop a robust email classification system to differentiate between legitimate emails and spam, thus offering users a clutter-free and secure communication experience. The project was successful in achieving this goal through a comprehensive approach, including data preprocessing, feature engineering, model selection, and evaluation.\n",
        "\n",
        "The first step involved transforming raw email text into a structured format compatible with machine learning algorithms. This was achieved using the Bag of Words (BoW) method, which translates textual data into numerical vectors. The data was then preprocessed and cleaned, with careful attention to HTML tags, special characters, and email metadata. The cleaned data was found to be 99% accurate.\n",
        "\n",
        "In the subsequent phase, the data was clustered to group similar emails based on their content. This innovative approach aimed to uncover latent patterns and structures within the email data. The clustering process revealed that text/html and text/plain constituted a significant 92% of the dataset, with the remaining 8% comprising various multipart messages.\n",
        "\n",
        "The main data modeling approach employed in the project was the Gaussian Naive Bayes (NB) classifier, which proved effective for the text classification task. The model was trained using a combination of TF-IDF features and cluster labels, and its performance was assessed using a 5-fold cross-validation technique. The model achieved impressive results, with an overall accuracy of 98% and high precision and recall scores for both spam and non-spam emails.\n",
        "\n",
        "The project also explored threshold tuning to enhance the model's performance, especially in improving spam email recall. By adjusting the threshold, the model could be optimized to reduce false positives or false negatives, depending on the goal.\n",
        "\n",
        "In conclusion, the project successfully developed a spam classification system that leverages machine learning and natural language processing techniques to classify emails accurately. This system has significant implications for email filtering, cybersecurity, and information management, contributing to a safer and more efficient digital communication environment."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "machine_shape": "hm",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "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": "3.11.4"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}