spam filter.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                    MAIL   fil\n",
      "0  send send us send us us your password  spam\n",
      "1                    send us your review   ham\n",
      "2                   review your password   ham\n",
      "3                              review us  spam\n",
      "4                  send us your password  spam\n",
      "5                   send us your account  spam\n",
      "{'send': 6, 'us': 7, 'your': 5, 'password': 3, 'review': 3, 'account': 1}\n",
      "{'spam': 4, 'ham': 2}\n",
      "Total words:  25\n",
      "{'spam': {'send': 0.75, 'us': 1.0, 'your': 0.75, 'password': 0.5, 'review': 0.25, 'account': 0.25}, 'ham': {'send': 0.5, 'us': 0.5, 'your': 1.0, 'review': 1.0, 'password': 0.5}}\n",
      "{'spam': 0.6666666666666666, 'ham': 0.3333333333333333}\n",
      "0.75\n",
      "0.3333333333333333\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "data = pd.read_csv(\"spam.csv\")\n",
    "print(data)\n",
    "\n",
    "bow = {}\n",
    "mails = data[\"MAIL\"]\n",
    "tot_word_count = 0\n",
    "\n",
    "for mail in mails:\n",
    "    mail = mail.split()\n",
    "    for word in mail:\n",
    "        if word in bow:\n",
    "            bow[word] += 1\n",
    "            tot_word_count += 1\n",
    "        else:\n",
    "            bow[word] = 1\n",
    "            tot_word_count += 1\n",
    "            \n",
    "print(bow)\n",
    "\n",
    "fil = {'spam': 0, 'ham': 0}\n",
    "for fils in data[\"fil\"]:\n",
    "    fil[fils] += 1\n",
    "print(fil)\n",
    "print(\"Total words: \", tot_word_count)\n",
    "\n",
    "prob = {'spam': {}, 'ham': {}}\n",
    "for index, row in data.iterrows():\n",
    "    mail = row['MAIL']\n",
    "    res = row['fil']\n",
    "    done = []\n",
    "    for word in mail.split():\n",
    "        if word not in done:\n",
    "            done.append(word)\n",
    "            if word not in prob[res]:\n",
    "                prob[res][word] = 1\n",
    "            else:\n",
    "                prob[res][word] += 1\n",
    "target_prob = {'spam': fil['spam']/len(data), 'ham': fil['ham']/len(data)}\n",
    "\n",
    "for res in prob:\n",
    "    for word in prob[res]:\n",
    "        prob[res][word] = prob[res][word] / fil[res]\n",
    "print(prob)\n",
    "print(target_prob)\n",
    "\n",
    "word = \"send\"\n",
    "\n",
    "p_ws = prob['spam'].get(word, 0)\n",
    "p_s = target_prob['spam']\n",
    "p_wh = prob['ham'].get(word, 0)\n",
    "p_h = target_prob['ham']\n",
    "\n",
    "p_sw = p_ws * p_s / (p_ws * p_s + p_wh * p_h)\n",
    "print(p_sw)\n",
    "\n",
    "word1 = \"review\"\n",
    "word2 = \"password\"\n",
    "\n",
    "p_w1s = prob['spam'].get(word1, 0)\n",
    "p_s = target_prob['spam']\n",
    "p_w1h = prob['ham'].get(word1, 0)\n",
    "p_h = target_prob['ham']\n",
    "\n",
    "p_w2s = prob['spam'].get(word2, 0)\n",
    "p_w2h = prob['ham'].get(word2, 0)\n",
    "\n",
    "p_sww = p_w1s * p_w2s * p_s / (p_w1s * p_w2s * p_s + p_w1h * p_w2h * p_h)\n",
    "print(p_sww)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}