{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# IndoHoaxDetector Example Notebook\n",
    "\n",
    "This notebook demonstrates how to use the IndoHoaxDetector model to classify Indonesian news articles as hoax or legitimate.\n",
    "\n",
    "## Setup\n",
    "\n",
    "First, make sure you have the required libraries installed:\n",
    "```bash\n",
    "pip install scikit-learn\n",
    "```\n",
    "\n",
    "And ensure the `logreg_model.pkl` file is in the same directory as this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import numpy as np\n",
    "\n",
    "# Load the model\n",
    "with open('logreg_model.pkl', 'rb') as f:\n",
    "    model = pickle.load(f)\n",
    "\n",
    "print(\"Model loaded successfully!\")\n",
    "print(f\"Model type: {type(model)}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Example Predictions\n",
    "\n",
    "Let's test the model with some example Indonesian news texts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Example texts\n",
    "test_texts = [\n",
    "    \"Presiden Joko Widodo mengumumkan program vaksinasi COVID-19 untuk seluruh masyarakat Indonesia.\",\n",
    "    \"Ditemukan dinosaurus hidup di Danau Toba, Sumatera Utara.\",\n",
    "    \"Harga BBM akan naik mulai bulan depan akibat kenaikan harga minyak dunia.\",\n",
    "    \"Minum jus lemon setiap hari bisa menyembuhkan diabetes secara permanen.\"\n",
    "]\n",
    "\n",
    "# Make predictions\n",
    "predictions = model.predict(test_texts)\n",
    "probabilities = model.predict_proba(test_texts)\n",
    "\n",
    "# Display results\n",
    "for i, (text, pred, prob) in enumerate(zip(test_texts, predictions, probabilities)):\n",
    "    label = \"Hoax\" if pred == 1 else \"Legitimate\"\n",
    "    confidence = prob[pred]\n",
    "    print(f\"Example {i+1}:\")\n",
    "    print(f\"Text: {text[:80]}{'...' if len(text) > 80 else ''}\")\n",
    "    print(f\"Prediction: {label} (Confidence: {confidence:.4f})\")\n",
    "    print(f\"Probabilities: Legitimate={prob[0]:.4f}, Hoax={prob[1]:.4f}\")\n",
    "    print(\"-\" * 50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Custom Prediction\n",
    "\n",
    "Try your own text below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Enter your own Indonesian news text here\n",
    "custom_text = \"Masukkan berita Indonesia yang ingin Anda periksa di sini.\"\n",
    "\n",
    "# Make prediction\n",
    "custom_pred = model.predict([custom_text])[0]\n",
    "custom_prob = model.predict_proba([custom_text])[0]\n",
    "\n",
    "custom_label = \"Hoax\" if custom_pred == 1 else \"Legitimate\"\n",
    "custom_confidence = custom_prob[custom_pred]\n",
    "\n",
    "print(f\"Custom Text: {custom_text}\")\n",
    "print(f\"Prediction: {custom_label}\")\n",
    "print(f\"Confidence: {custom_confidence:.4f}\")\n",
    "print(f\"Probabilities: Legitimate={custom_prob[0]:.4f}, Hoax={custom_prob[1]:.4f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Notes\n",
    "\n",
    "- This model is trained on Indonesian news data and may not work well with other languages.\n",
    "- Always use human judgment and verify information from multiple sources.\n",
    "- The model provides probability scores that can help assess confidence in predictions."
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}