project03_NBLR.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "project03_NBLR.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JBXZCR1b0YPe"
      },
      "source": [
        "# Import and Reading Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_mb6iuKtQk66"
      },
      "source": [
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer \n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.pipeline import make_pipeline "
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "12V-PqUFb7FL",
        "outputId": "93473212-599f-4d4b-bbb8-94b9a083eef7"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vQ9ilrMeP6sw",
        "outputId": "4ec37c51-0497-4840-b792-3b5ec38c5081"
      },
      "source": [
        "test_path = \"/content/drive/MyDrive/Colab Notebooks/nlp-project-3-dataset/test.csv\"\n",
        "train_path = \"/content/drive/MyDrive/Colab Notebooks/nlp-project-3-dataset/train.csv\"\n",
        "import nltk\n",
        "nltk.download('stopwords')\n",
        "stop_word_list = nltk.corpus.stopwords.words('turkish')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JeHg3R1h0C1_"
      },
      "source": [
        "## Reading Files"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OE7hwg_OQ3Tu"
      },
      "source": [
        "test_df = pd.read_csv(test_path)\n",
        "train_df = pd.read_csv(train_path)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "id": "WB3RNI4ESZIH",
        "outputId": "22216dee-1f3a-4494-c3fa-8f4520c73540"
      },
      "source": [
        "train_df.head(10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Unnamed: 0</th>\n",
              "      <th>id</th>\n",
              "      <th>text</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>547530</td>\n",
              "      <td>Beşiktaş'ın eski teknik direktörü Slaven Bilic...</td>\n",
              "      <td>spor</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>42683</td>\n",
              "      <td>14 Şubat, Katolik Kilisesi’nin azizlerinden St...</td>\n",
              "      <td>yazarlar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>2</td>\n",
              "      <td>770925</td>\n",
              "      <td>İstanbul Kağıthane’de gece yarısı sokakta oyun...</td>\n",
              "      <td>video</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>473224</td>\n",
              "      <td>Ziraat Türkiye Kupası maçında deplasmanda Amed...</td>\n",
              "      <td>video</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>4</td>\n",
              "      <td>44833</td>\n",
              "      <td>Eski Brezilyalı yıldız oyuncu Ronaldo, Brezily...</td>\n",
              "      <td>spor</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>5</td>\n",
              "      <td>343736</td>\n",
              "      <td>\\n\\n\\n \\nLideri, anneler gününü Hazreti Muhamm...</td>\n",
              "      <td>yazarlar</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>6</td>\n",
              "      <td>729677</td>\n",
              "      <td>İngiltere'nin Daily Star Gazetesinin son dakik...</td>\n",
              "      <td>spor</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>7</td>\n",
              "      <td>37307</td>\n",
              "      <td>Irak’ın Nineva eyaletinde bir Irak Şam İslam D...</td>\n",
              "      <td>dunya</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>8</td>\n",
              "      <td>278921</td>\n",
              "      <td>Meteoroloji'den son dakika hava durumu tahmini...</td>\n",
              "      <td>turkiye</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>9</td>\n",
              "      <td>554302</td>\n",
              "      <td>Avrupa Futbol Şampiyonası'nda (EURO 2016) gols...</td>\n",
              "      <td>spor</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Unnamed: 0  ...     label\n",
              "0           0  ...      spor\n",
              "1           1  ...  yazarlar\n",
              "2           2  ...     video\n",
              "3           3  ...     video\n",
              "4           4  ...      spor\n",
              "5           5  ...  yazarlar\n",
              "6           6  ...      spor\n",
              "7           7  ...     dunya\n",
              "8           8  ...   turkiye\n",
              "9           9  ...      spor\n",
              "\n",
              "[10 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dMzs8T6J0HU7"
      },
      "source": [
        "## Dividing data into keys and values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZIWzISPQRFRx"
      },
      "source": [
        "testx = test_df[\"text\"]\n",
        "testy = test_df[\"label\"]\n",
        "trainx = train_df[\"text\"]\n",
        "trainy = train_df[\"label\"]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "M-_FlGXyS4yg",
        "outputId": "8bed76d1-cf9a-49df-91d5-58b2a603ee64"
      },
      "source": [
        "print(\"train size:\",trainx.size)\n",
        "print(\"test.size:\",testx.size)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "train size: 8000\n",
            "test.size: 2000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vVjclgK20Stw"
      },
      "source": [
        "# Naïve Bayes"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "80pLawCDTeb7"
      },
      "source": [
        "from sklearn.pipeline import Pipeline\n",
        "model_pipeline=Pipeline([('tfidf_vectorizer',TfidfVectorizer(lowercase=True, stop_words=stop_word_list,)),(\"mnb\",MultinomialNB())]) # extracted stopwords and lowercased"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "u1LBQILVfOpV",
        "outputId": "34209fb5-c109-4d20-d1ad-fdfb8e648373"
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "\n",
        "\n",
        "grid_params = {\n",
        "  'mnb__alpha': np.linspace(0.5, 1.5, 3),\n",
        "  'mnb__fit_prior': [True, False],\n",
        "  'tfidf_vectorizer__use_idf':[True,False], # include idf or not\n",
        "  'tfidf_vectorizer__norm': [None, 'l1', 'l2'], # normalization\n",
        "  'tfidf_vectorizer__binary': [True, False] \n",
        "}\n",
        "clf = GridSearchCV(model_pipeline, grid_params,verbose=10,n_jobs=-1)\n",
        "clf.fit(trainx, trainy)\n",
        "print(\"Best Score: \", clf.best_score_)\n",
        "print(\"Best Params: \", clf.best_params_)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fitting 5 folds for each of 72 candidates, totalling 360 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 2 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done   1 tasks      | elapsed:    4.8s\n",
            "[Parallel(n_jobs=-1)]: Done   4 tasks      | elapsed:    8.5s\n",
            "[Parallel(n_jobs=-1)]: Done   9 tasks      | elapsed:   19.0s\n",
            "[Parallel(n_jobs=-1)]: Done  14 tasks      | elapsed:   26.4s\n",
            "[Parallel(n_jobs=-1)]: Done  21 tasks      | elapsed:   40.4s\n",
            "[Parallel(n_jobs=-1)]: Done  28 tasks      | elapsed:   51.2s\n",
            "[Parallel(n_jobs=-1)]: Done  37 tasks      | elapsed:  1.1min\n",
            "[Parallel(n_jobs=-1)]: Done  46 tasks      | elapsed:  1.4min\n",
            "[Parallel(n_jobs=-1)]: Done  57 tasks      | elapsed:  1.7min\n",
            "[Parallel(n_jobs=-1)]: Done  68 tasks      | elapsed:  2.0min\n",
            "[Parallel(n_jobs=-1)]: Done  81 tasks      | elapsed:  2.4min\n",
            "[Parallel(n_jobs=-1)]: Done  94 tasks      | elapsed:  2.8min\n",
            "[Parallel(n_jobs=-1)]: Done 109 tasks      | elapsed:  3.3min\n",
            "[Parallel(n_jobs=-1)]: Done 124 tasks      | elapsed:  3.7min\n",
            "[Parallel(n_jobs=-1)]: Done 141 tasks      | elapsed:  4.2min\n",
            "[Parallel(n_jobs=-1)]: Done 158 tasks      | elapsed:  4.7min\n",
            "[Parallel(n_jobs=-1)]: Done 177 tasks      | elapsed:  5.3min\n",
            "[Parallel(n_jobs=-1)]: Done 196 tasks      | elapsed:  5.8min\n",
            "[Parallel(n_jobs=-1)]: Done 217 tasks      | elapsed:  6.5min\n",
            "[Parallel(n_jobs=-1)]: Done 238 tasks      | elapsed:  7.1min\n",
            "[Parallel(n_jobs=-1)]: Done 261 tasks      | elapsed:  7.7min\n",
            "[Parallel(n_jobs=-1)]: Done 284 tasks      | elapsed:  8.4min\n",
            "[Parallel(n_jobs=-1)]: Done 309 tasks      | elapsed:  9.1min\n",
            "[Parallel(n_jobs=-1)]: Done 334 tasks      | elapsed:  9.9min\n",
            "[Parallel(n_jobs=-1)]: Done 360 out of 360 | elapsed: 10.6min finished\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Best Score:  0.789125\n",
            "Best Params:  {'mnb__alpha': 1.5, 'mnb__fit_prior': True, 'tfidf_vectorizer__binary': True, 'tfidf_vectorizer__norm': None, 'tfidf_vectorizer__use_idf': True}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-aHANaYJ5GzG",
        "outputId": "0a05148a-94d6-44c4-fd2e-0880ee48f909"
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "y_pred = clf.predict(testx)\n",
        "print(classification_report(testy, y_pred))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "       dunya       0.80      0.85      0.82       395\n",
            "        spor       0.90      0.93      0.92       384\n",
            "     turkiye       0.65      0.67      0.66       421\n",
            "       video       0.74      0.51      0.60       408\n",
            "    yazarlar       0.80      0.96      0.87       392\n",
            "\n",
            "    accuracy                           0.78      2000\n",
            "   macro avg       0.78      0.78      0.78      2000\n",
            "weighted avg       0.78      0.78      0.77      2000\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        },
        "id": "CvzU93JXeudr",
        "outputId": "1dc3aff9-b105-4f62-dfd8-f08cf3b86501"
      },
      "source": [
        "# from sklearn.metrics import confusion_matrix\n",
        "# print(confusion_matrix(testy, y_pred))\n",
        "\n",
        "\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.metrics import plot_confusion_matrix\n",
        "\n",
        "# Run classifier, using a model that is too regularized (C too low) to see\n",
        "# the impact on the results\n",
        "\n",
        "# np.set_printoptions(precision=5)\n",
        "\n",
        "class_names = ['spor','yazarlar','video','dunya','turkiye']\n",
        "\n",
        "# Plot non-normalized confusion matrix\n",
        "titles_options = \"Naive Bayes Confusion matrix\"\n",
        "disp = plot_confusion_matrix(clf, testx, testy,\n",
        "                              # display_labels=class_names,\n",
        "                              cmap=plt.cm.Blues,normalize=\"true\")\n",
        "disp.ax_.set_title(titles_options)\n",
        "\n",
        "\n",
        "plt.show()\n",
        "plt.clf()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2rhlk6Y900_d"
      },
      "source": [
        "# Logistic Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "OBzieNeoArcq",
        "outputId": "bb286e39-11b5-40f4-b58f-65501d15a7e3"
      },
      "source": [
        "import numpy as np\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "\n",
        "lr_pipeline = Pipeline([\n",
        "    ('tfidf_vectorizer', TfidfVectorizer(lowercase=True, stop_words=stop_word_list)), # extracted stopwords and lowercased\n",
        "    ('lr', LogisticRegression(max_iter=1000))\n",
        "])\n",
        "\n",
        "\n",
        "grid_params = {\n",
        "  'lr__C': np.logspace(-3,3,3),\n",
        "  'tfidf_vectorizer__use_idf':[True,False],\n",
        "  'tfidf_vectorizer__binary': [True, False]\n",
        "}\n",
        "\n",
        "clf = GridSearchCV(lr_pipeline, grid_params,verbose=5,n_jobs=-1)\n",
        "clf.fit(trainx, trainy)\n",
        "\n",
        "print(\"Best Score: \", clf.best_score_)\n",
        "print(\"Best Params: \", clf.best_params_)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Fitting 5 folds for each of 12 candidates, totalling 60 fits\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 2 concurrent workers.\n",
            "[Parallel(n_jobs=-1)]: Done  14 tasks      | elapsed:   33.3s\n",
            "[Parallel(n_jobs=-1)]: Done  60 out of  60 | elapsed: 16.0min finished\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Best Score:  0.8564999999999999\n",
            "Best Params:  {'lr__C': 1000.0, 'tfidf_vectorizer__binary': True, 'tfidf_vectorizer__use_idf': True}\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8zJ7mx3VjDqJ",
        "outputId": "9ecdf50b-754e-4011-a39a-5476b5b89afc"
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "y_pred = clf.predict(testx)\n",
        "print(classification_report(testy, y_pred))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "       dunya       0.82      0.88      0.85       395\n",
            "        spor       0.93      0.94      0.94       384\n",
            "     turkiye       0.82      0.73      0.77       421\n",
            "       video       0.81      0.80      0.81       408\n",
            "    yazarlar       0.91      0.96      0.93       392\n",
            "\n",
            "    accuracy                           0.86      2000\n",
            "   macro avg       0.86      0.86      0.86      2000\n",
            "weighted avg       0.86      0.86      0.86      2000\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        },
        "id": "T7MT78kbjFYM",
        "outputId": "10a92020-3176-449a-d6c1-091296b0c72c"
      },
      "source": [
        "class_names = ['spor','yazarlar','video','dunya','turkiye']\n",
        "\n",
        "# Plot non-normalized confusion matrix\n",
        "titles_options = \"Logistic Regression Confusion matrix\"\n",
        "disp = plot_confusion_matrix(clf, testx, testy,\n",
        "                              # display_labels=class_names,\n",
        "                              cmap=plt.cm.Blues,normalize=\"true\")\n",
        "disp.ax_.set_title(titles_options)\n",
        "\n",
        "\n",
        "plt.show()\n",
        "plt.clf()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}