spam_detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8gEVQHSCP4sC"
   },
   "source": [
    "### IMPORTS\n",
    "\n",
    "---\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "executionInfo": {
     "elapsed": 350,
     "status": "ok",
     "timestamp": 1644925164796,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "Vd8Xv-bbP2l3"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7O0EOPeUQd4R"
   },
   "source": [
    "### DATASET\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "executionInfo": {
     "elapsed": 224,
     "status": "ok",
     "timestamp": 1644925166187,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "5PJMY-nVXPwZ"
   },
   "outputs": [],
   "source": [
    "dataset = pd.read_csv(\"spam.csv\",\n",
    "                      encoding=\"latin1\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1644925167424,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "Ad35R0yJ5Usk",
    "outputId": "e047bb05-3d93-4a62-98c3-6683b57ad276"
   },
   "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>v1</th>\n",
       "      <th>v2</th>\n",
       "      <th>Unnamed: 2</th>\n",
       "      <th>Unnamed: 3</th>\n",
       "      <th>Unnamed: 4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ham</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ham</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spam</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ham</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ham</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     v1                                                 v2 Unnamed: 2  \\\n",
       "0   ham  Go until jurong point, crazy.. Available only ...        NaN   \n",
       "1   ham                      Ok lar... Joking wif u oni...        NaN   \n",
       "2  spam  Free entry in 2 a wkly comp to win FA Cup fina...        NaN   \n",
       "3   ham  U dun say so early hor... U c already then say...        NaN   \n",
       "4   ham  Nah I don't think he goes to usf, he lives aro...        NaN   \n",
       "\n",
       "  Unnamed: 3 Unnamed: 4  \n",
       "0        NaN        NaN  \n",
       "1        NaN        NaN  \n",
       "2        NaN        NaN  \n",
       "3        NaN        NaN  \n",
       "4        NaN        NaN  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "executionInfo": {
     "elapsed": 327,
     "status": "ok",
     "timestamp": 1644925168627,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "cVd2Y0xK8gDz"
   },
   "outputs": [],
   "source": [
    "# get rid of unneccessary columns\n",
    "dataset.drop([\"Unnamed: 2\", \"Unnamed: 3\", \"Unnamed: 4\"],\n",
    "          axis=1,\n",
    "          inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "executionInfo": {
     "elapsed": 309,
     "status": "ok",
     "timestamp": 1644925170048,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "8Llqcbe88C6z"
   },
   "outputs": [],
   "source": [
    "dataset.columns = [\"labels\", \"text\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1644925170984,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "htknDT1q8MgK",
    "outputId": "4e6a344c-68c1-4400-8a5f-0e1f97f76256"
   },
   "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>labels</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ham</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ham</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spam</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ham</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ham</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  labels                                               text\n",
       "0    ham  Go until jurong point, crazy.. Available only ...\n",
       "1    ham                      Ok lar... Joking wif u oni...\n",
       "2   spam  Free entry in 2 a wkly comp to win FA Cup fina...\n",
       "3    ham  U dun say so early hor... U c already then say...\n",
       "4    ham  Nah I don't think he goes to usf, he lives aro..."
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644925172073,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "OCdl6JcYio8v"
   },
   "outputs": [],
   "source": [
    "# change labels\n",
    "data_labels = dataset[\"labels\"].values\n",
    "data_labels[data_labels == \"spam\"] = 1\n",
    "data_labels[data_labels == \"ham\"] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "executionInfo": {
     "elapsed": 222,
     "status": "ok",
     "timestamp": 1644925173230,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "SknRLf_D6GNk",
    "outputId": "5db1cc7f-20cb-46b7-9d91-a1d385d1e5ce"
   },
   "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>labels</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  labels                                               text\n",
       "0      0  Go until jurong point, crazy.. Available only ...\n",
       "1      0                      Ok lar... Joking wif u oni...\n",
       "2      1  Free entry in 2 a wkly comp to win FA Cup fina...\n",
       "3      0  U dun say so early hor... U c already then say...\n",
       "4      0  Nah I don't think he goes to usf, he lives aro..."
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 344,
     "status": "ok",
     "timestamp": 1644925174814,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "wgbabJCG6mw-",
    "outputId": "f75b3079-edf5-41f8-a40f-db497d4871c6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    4825\n",
       "1     747\n",
       "Name: labels, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset[\"labels\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 427
    },
    "executionInfo": {
     "elapsed": 516,
     "status": "ok",
     "timestamp": 1644925176260,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "L2yb3X2Jnir0",
    "outputId": "8f33e163-9dcd-47a2-ae90-e0ae55924754"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# check if the dataset is balanced\n",
    "labels = [\"not_spam\", \"spam\"]\n",
    "colors = [\"olivedrab\", \"silver\"]\n",
    "\n",
    "plt.figure(figsize=(7,7))\n",
    "plt.pie(dataset[\"labels\"].value_counts(), labels=labels, colors=colors, autopct=\"%1.1f%%\")\n",
    "plt.legend(title = \"Spam status\")\n",
    "plt.title(\"Dataset distribution\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B5b8PhnMQutp"
   },
   "source": [
    "### DATA PREPROCESSING\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "executionInfo": {
     "elapsed": 215,
     "status": "ok",
     "timestamp": 1644925799540,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "UMw9InhL99h5"
   },
   "outputs": [],
   "source": [
    "X = dataset[\"text\"]\n",
    "y = dataset[\"labels\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 224,
     "status": "ok",
     "timestamp": 1644925800986,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "A8RWmBn__HgB",
    "outputId": "01844495-842b-46a4-8430-fcc93408974e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5572,)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644925801814,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "eHsB1Yfw_IoB",
    "outputId": "abd5cbc8-b776-4101-b001-d6c0046a6e12"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5572,)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Jwmp_9FJFahE"
   },
   "source": [
    "**EMBEDDING**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "executionInfo": {
     "elapsed": 208,
     "status": "ok",
     "timestamp": 1644925803366,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "yQYogq_j-SwT"
   },
   "outputs": [],
   "source": [
    "from keras.preprocessing.text import Tokenizer\n",
    "\n",
    "tokenizer = Tokenizer()\n",
    "tokenizer.fit_on_texts(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644925804408,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "4J3PiQXU-dSj"
   },
   "outputs": [],
   "source": [
    "embedded_data = tokenizer.texts_to_sequences(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CFu7eiopFc3z"
   },
   "source": [
    "**PADDING**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "executionInfo": {
     "elapsed": 1,
     "status": "ok",
     "timestamp": 1644925805533,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "j9_ACPZ79jyS"
   },
   "outputs": [],
   "source": [
    "from keras.preprocessing.sequence import pad_sequences\n",
    "\n",
    "# pad to the longest item in the dataset\n",
    "max_length = max([len(text) for text in X])\n",
    "\n",
    "X = pad_sequences(embedded_data,\n",
    "                  value=0,\n",
    "                  padding=\"post\",\n",
    "                  maxlen=max_length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644925806345,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "JI3c0Vo59nmS",
    "outputId": "2fe81320-0325-4943-caf7-4e4feff8b7ae"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  50,  469, 4410, ...,    0,    0,    0],\n",
       "       [  46,  336, 1495, ...,    0,    0,    0],\n",
       "       [  47,  486,    8, ...,    0,    0,    0],\n",
       "       [   6,  245,  152, ...,    0,    0,    0],\n",
       "       [1018,    1,   98, ...,    0,    0,    0]])"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5uX1jZHpQ3AJ"
   },
   "source": [
    "### TRAIN/TEST/VALIDATION SPLIT\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "executionInfo": {
     "elapsed": 235,
     "status": "ok",
     "timestamp": 1644934883517,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "OVMzfOdA9IGK"
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_remaining, y_train, y_remaining = train_test_split(X, y, test_size = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "executionInfo": {
     "elapsed": 1,
     "status": "ok",
     "timestamp": 1644934884532,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "84vRDmwgDV82"
   },
   "outputs": [],
   "source": [
    "# second split for the validation set\n",
    "X_valid, X_test, y_valid, y_test = train_test_split(X_remaining, y_remaining, test_size = 0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644934325224,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "r1Hnb-tf_WTZ",
    "outputId": "a76f773f-1a58-4d7d-a579-81f8972051e1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  30,   25,  189, ...,    0,    0,    0],\n",
       "       [ 161,   39, 1921, ...,    0,    0,    0],\n",
       "       [ 161,   31,    5, ...,    0,    0,    0],\n",
       "       [ 720,  707,  227, ...,    0,    0,    0],\n",
       "       [  13,  330,  119, ...,    0,    0,    0]])"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644934326231,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "8tkmE7AaBJxQ",
    "outputId": "6d8cfb14-b6f7-434c-e38a-048842313d35"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[   5, 6660,  477, ...,    0,    0,    0],\n",
       "       [   1,  136,  916, ...,    0,    0,    0],\n",
       "       [  74,   15,  288, ...,    0,    0,    0],\n",
       "       [1837,  355,   21, ...,    0,    0,    0],\n",
       "       [2847,    7, 1781, ...,    0,    0,    0]])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1644934327202,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "SG-4XYJpDczX",
    "outputId": "966e8a07-59ad-423d-fd61-8034254247d3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 45,  22,   3, ...,   0,   0,   0],\n",
       "       [ 10, 134, 372, ...,   0,   0,   0],\n",
       "       [ 28,   1, 106, ...,   0,   0,   0],\n",
       "       [ 53,   6, 875, ...,   0,   0,   0],\n",
       "       [123,  83, 106, ...,   0,   0,   0]])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_valid[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644934328253,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "otuzkqGBDdEu",
    "outputId": "6a205c74-088d-4c84-d360-0fc5b1958431"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4505    0\n",
       "1079    0\n",
       "2170    0\n",
       "3060    0\n",
       "2037    0\n",
       "Name: labels, dtype: object"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_valid[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1644934329235,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "UipkHjXO_XJ5",
    "outputId": "25792eb1-6dab-4a81-9ed2-7d13bb8becc1"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4767    0\n",
       "3398    0\n",
       "2785    0\n",
       "414     1\n",
       "1466    0\n",
       "Name: labels, dtype: object"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 222,
     "status": "ok",
     "timestamp": 1644934330912,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "lHvhNEglBGFA",
    "outputId": "51c98961-d31a-4b2e-a7b7-f5e1507f0e9a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2793    0\n",
       "1973    0\n",
       "2406    0\n",
       "1705    0\n",
       "5517    0\n",
       "Name: labels, dtype: object"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GjndbhOQSHwP"
   },
   "source": [
    "### RANDOM FOREST"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Kam4ZR5LScp-"
   },
   "source": [
    "**MODEL CONFIGURATION**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "executionInfo": {
     "elapsed": 219,
     "status": "ok",
     "timestamp": 1644935278678,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "1UuPN4HIAIc3"
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "RF_classifier = RandomForestClassifier(n_estimators = 10, criterion = 'entropy')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eF5D8N0fSfRW"
   },
   "source": [
    "**MODEL TRAINING**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# fixes some errors\n",
    "y_train = np.array(y_train, dtype=float)\n",
    "y_valid = np.array(y_valid, dtype=float)\n",
    "y_test = np.array(y_test, dtype=float)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 211,
     "status": "ok",
     "timestamp": 1644935280212,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "4n-2LUx7AJwM",
    "outputId": "143ccedf-292b-4930-ea4d-83396e0c396f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(criterion='entropy', n_estimators=10)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "RF_classifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uPrgRPQTShm3"
   },
   "source": [
    "**MODEL EVALUATION**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "executionInfo": {
     "elapsed": 254,
     "status": "ok",
     "timestamp": 1644935283274,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "i5QWtNLcAKFo"
   },
   "outputs": [],
   "source": [
    "y_pred = RF_classifier.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644935284571,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "5TkqXLry4tKH",
    "outputId": "d8f91fec-691e-4acb-c8eb-7a0f4070585a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.94      0.99      0.96       488\n",
      "         1.0       0.89      0.56      0.68        70\n",
      "\n",
      "    accuracy                           0.94       558\n",
      "   macro avg       0.91      0.77      0.82       558\n",
      "weighted avg       0.93      0.94      0.93       558\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ppt2b8CiSjWu"
   },
   "source": [
    "**NEW PREDICTIONS**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "executionInfo": {
     "elapsed": 248,
     "status": "ok",
     "timestamp": 1644935430717,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "iEHK5PqjEO7u"
   },
   "outputs": [],
   "source": [
    "my_message = [\"hi r u lookin 4 saucy fun wiv busty woman am free sex chat now +491574546387\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "executionInfo": {
     "elapsed": 1,
     "status": "ok",
     "timestamp": 1644935431753,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "EDo94TIBCSeH"
   },
   "outputs": [],
   "source": [
    "def preprocess(text):\n",
    "  text = tokenizer.texts_to_sequences(text)\n",
    "  text = pad_sequences(text,\n",
    "                      value=0,\n",
    "                      padding=\"post\",\n",
    "                      maxlen=max_length)\n",
    "  return text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "executionInfo": {
     "elapsed": 1,
     "status": "ok",
     "timestamp": 1644935432860,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "qE7XjgpFJEa8"
   },
   "outputs": [],
   "source": [
    "new_message = preprocess(my_message)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644935433706,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "5Mg5_K79AKoG"
   },
   "outputs": [],
   "source": [
    "output = RF_classifier.predict(new_message)\n",
    "\n",
    "def prediction(output):\n",
    "   if output[0] == 1:\n",
    "     return \"spam\"\n",
    "   else:\n",
    "     return \"not_spam\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "executionInfo": {
     "elapsed": 3,
     "status": "ok",
     "timestamp": 1644935434514,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "iDqxZe7CkFnK",
    "outputId": "ad57033e-af06-46d6-e75f-f004e202eff3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'not_spam'"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# so wrong\n",
    "prediction(output)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "p6wEhMJIRPxw"
   },
   "source": [
    "### RECURRENT NEURAL NETWORK"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "KsfmMGbsQ5JZ"
   },
   "source": [
    "**MODEL CONFIGURATION**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "executionInfo": {
     "elapsed": 253,
     "status": "ok",
     "timestamp": 1644932151044,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "SQb4UkBp26SJ"
   },
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Input, Embedding, LSTM, GlobalMaxPool1D, Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "executionInfo": {
     "elapsed": 229,
     "status": "ok",
     "timestamp": 1644934357199,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "CZ1lUGwRSXMP"
   },
   "outputs": [],
   "source": [
    "INPUT_SHAPE = X_train.shape[1]\n",
    "VOCABULARY_SIZE = len(tokenizer.word_index)\n",
    "EMBEDDING_DIMENSION = 20\n",
    "LSTM_UNITS = 15"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "executionInfo": {
     "elapsed": 1142,
     "status": "ok",
     "timestamp": 1644934441042,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "zWkkHguwHCvV"
   },
   "outputs": [],
   "source": [
    "rnn = Sequential()\n",
    "\n",
    "rnn.add(Input(INPUT_SHAPE,))\n",
    "# +1 here resolves the indexing problem during training\n",
    "rnn.add(Embedding(VOCABULARY_SIZE+1, EMBEDDING_DIMENSION))\n",
    "\n",
    "rnn.add(LSTM(LSTM_UNITS, return_sequences=True))\n",
    "\n",
    "rnn.add(GlobalMaxPool1D())\n",
    "\n",
    "rnn.add(Dense(units=1, activation=\"sigmoid\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "executionInfo": {
     "elapsed": 228,
     "status": "ok",
     "timestamp": 1644934442636,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "j4fZw47B8Z3R"
   },
   "outputs": [],
   "source": [
    "rnn.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 355,
     "status": "ok",
     "timestamp": 1644934443826,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "BzaByCkK8aJx",
    "outputId": "88adceaf-001d-4cf2-ffc4-a130773c57ef"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " embedding (Embedding)       (None, 910, 20)           178420    \n",
      "                                                                 \n",
      " lstm (LSTM)                 (None, 910, 15)           2160      \n",
      "                                                                 \n",
      " global_max_pooling1d (Globa  (None, 15)               0         \n",
      " lMaxPooling1D)                                                  \n",
      "                                                                 \n",
      " dense (Dense)               (None, 1)                 16        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 180,596\n",
      "Trainable params: 180,596\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "rnn.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YR78AqTJQ9fp"
   },
   "source": [
    "**MODEL TRAINING**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "executionInfo": {
     "elapsed": 222,
     "status": "ok",
     "timestamp": 1644934445508,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "uarvoyr-qL4U"
   },
   "outputs": [],
   "source": [
    "BATCH_SIZE = 32\n",
    "EPOCHS = 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "id": "YPfcwyxQ81MJ",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "140/140 [==============================] - 16s 110ms/step - loss: 0.4911 - accuracy: 0.8452 - val_loss: 0.3281 - val_accuracy: 0.8564\n",
      "Epoch 2/5\n",
      "140/140 [==============================] - 15s 109ms/step - loss: 0.2266 - accuracy: 0.9226 - val_loss: 0.1455 - val_accuracy: 0.9677\n",
      "Epoch 3/5\n",
      "140/140 [==============================] - 15s 107ms/step - loss: 0.0942 - accuracy: 0.9856 - val_loss: 0.0905 - val_accuracy: 0.9749\n",
      "Epoch 4/5\n",
      "140/140 [==============================] - 15s 106ms/step - loss: 0.0508 - accuracy: 0.9942 - val_loss: 0.0770 - val_accuracy: 0.9749\n",
      "Epoch 5/5\n",
      "140/140 [==============================] - 15s 106ms/step - loss: 0.0309 - accuracy: 0.9955 - val_loss: 0.0683 - val_accuracy: 0.9767\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x23f5509dc40>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnn.fit(X_train, y_train, epochs=EPOCHS, validation_data=(X_valid, y_valid), batch_size=BATCH_SIZE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ImUqeNmvQ-y_"
   },
   "source": [
    "**MODEL EVALUATION**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 281
    },
    "executionInfo": {
     "elapsed": 696,
     "status": "ok",
     "timestamp": 1644934786260,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "n6ODifVfAT8B",
    "outputId": "50f48124-f5ad-4c83-91af-f473cc7347f1"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(rnn.history.history[\"loss\"], label=\"Loss\", color=colors[0])\n",
    "plt.plot(rnn.history.history[\"val_loss\"], label=\"Validation Loss\", color=colors[1])\n",
    "plt.title(\"Loss metrics\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 281
    },
    "executionInfo": {
     "elapsed": 763,
     "status": "ok",
     "timestamp": 1644934788947,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "WXFz3FEbASjn",
    "outputId": "5142fca0-10db-4e57-fecf-83cdc6b8f4d3"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(rnn.history.history[\"accuracy\"], label=\"Accuracy\", color=colors[0])\n",
    "plt.plot(rnn.history.history[\"val_accuracy\"], label=\"Validation accuracy\", color=colors[1])\n",
    "plt.title(\"Accuracy metrics\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18/18 - 0s - loss: 0.0537 - accuracy: 0.9857 - 429ms/epoch - 24ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.05373460426926613, 0.9856630563735962]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnn.evaluate(X_test, y_test, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rnn.save('saved_model/spam_model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#rnn = tf.keras.models.load_model('saved_model/spam_model')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rMS0NzQJRHIH"
   },
   "source": [
    "**NEW PREDICTIONS**\n",
    "\n",
    "---\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "executionInfo": {
     "elapsed": 224,
     "status": "ok",
     "timestamp": 1644935450257,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "r30wwtoM9YCJ"
   },
   "outputs": [],
   "source": [
    "output = rnn.predict(new_message)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "executionInfo": {
     "elapsed": 2,
     "status": "ok",
     "timestamp": 1644935451195,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "RLuXAsP49M-w"
   },
   "outputs": [],
   "source": [
    "def prediction(output):\n",
    "   if output[0][0] > 0.5:\n",
    "     return \"spam\"\n",
    "   else:\n",
    "     return \"not_spam\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1644935452017,
     "user": {
      "displayName": "Petra Bajac",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GisMWQ6Nwt62NYivub0ffOkvc3Im46ek_BgBPz3=s64",
      "userId": "17459354793322234878"
     },
     "user_tz": -60
    },
    "id": "dHvyLougD8Dc",
    "outputId": "6257a8a0-7518-446b-bd66-6f8129a23e27"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'spam'"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prediction(output)"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "spam_detection.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}