code.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "code.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DP3-V6eWXqoV",
        "outputId": "95efed3b-d402-4b0b-b941-7535e313faee"
      },
      "source": [
        "#mounting google drive\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6ADgiCgrXtLy",
        "outputId": "2114357a-5ce5-4185-8ec4-d81e00681982"
      },
      "source": [
        "import pandas as pd\n",
        "#opening the data file in read mode\n",
        "df = pd.read_csv('/content/drive/My Drive/nlp assignments/assignment 4/Sentiment Analysis Dataset.tsv', sep='\\t')\n",
        "print(df)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "            id  sentiment                                             review\n",
            "0       5814_8          1  With all this stuff going down at the moment w...\n",
            "1       2381_9          1  \\The Classic War of the Worlds\\\" by Timothy Hi...\n",
            "2       7759_3          0  The film starts with a manager (Nicholas Bell)...\n",
            "3       3630_4          0  It must be assumed that those who praised this...\n",
            "4       9495_8          1  Superbly trashy and wondrously unpretentious 8...\n",
            "...        ...        ...                                                ...\n",
            "24995   3453_3          0  It seems like more consideration has gone into...\n",
            "24996   5064_1          0  I don't believe they made this film. Completel...\n",
            "24997  10905_3          0  Guy is a loser. Can't get girls, needs to buil...\n",
            "24998  10194_3          0  This 30 minute documentary Buñuel made in the ...\n",
            "24999   8478_8          1  I saw this movie as a child and it broke my he...\n",
            "\n",
            "[25000 rows x 3 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ynk5CtWDbAfi"
      },
      "source": [
        "reviews=df.iloc[:,2].values\n",
        "labels=df.iloc[:,1].values\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "19GKV9hsnGt2"
      },
      "source": [
        "#parsing html\n",
        "from bs4 import BeautifulSoup\n",
        "import re\n",
        "\n",
        "def parseHtml(html):\n",
        "  soup = BeautifulSoup(html, 'html.parser')\n",
        "  return soup.get_text()\n",
        "\n",
        "def removeDigits(string):\n",
        "  for i in range(10):\n",
        "    string=string.replace(str(i),' ')\n",
        "  return string\n",
        "\n",
        "#removing html\n",
        "reviews=list(map(parseHtml, reviews))\n",
        "\n",
        "#removing digits\n",
        "reviews=list(map(removeDigits, reviews))\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TlJzPD7xROKl",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f365341a-ed67-48e9-a0bc-763e3db34255"
      },
      "source": [
        "#tokenizing\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "tokenizedText=[nltk.word_tokenize(item) for item in reviews]\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dyjJY4RMRJd1"
      },
      "source": [
        "#removing punctuation\n",
        "punc = '''!()-[]{};:'\"\\, <>./?@#$%^&*_~'''\n",
        "tokenizedText= [[word for word in review if word not in punc] for review in tokenizedText]\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8aCEDjcjMhmo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f8638505-93a0-44d9-b677-b384d3619017"
      },
      "source": [
        "import numpy as np\n",
        "#splitting the Dataset into train and test set\n",
        "totalRows=np.shape(tokenizedText)[0]\n",
        "\n",
        "splitRatio=0.75\n",
        "splitPoint=int(splitRatio*totalRows)\n",
        "\n",
        "trainReviews=tokenizedText[:splitPoint]\n",
        "trainLabels=labels[:splitPoint]\n",
        "testReviews=tokenizedText[splitPoint:]\n",
        "testLabels=labels[splitPoint:]\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/numpy/core/_asarray.py:83: VisibleDeprecationWarning: Creating an ndarray from ragged nested sequences (which is a list-or-tuple of lists-or-tuples-or ndarrays with different lengths or shapes) is deprecated. If you meant to do this, you must specify 'dtype=object' when creating the ndarray\n",
            "  return array(a, dtype, copy=False, order=order)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lejSbCrsBqXk"
      },
      "source": [
        "#learning word embeddings on training data using Gensim library\n",
        "from gensim.models import Word2Vec, KeyedVectors\n",
        "import nltk\n",
        "\n",
        "embeddingsSize=128\n",
        "model=Word2Vec(trainReviews, size=embeddingsSize, window=5, min_count=1, workers=4)\n",
        "######################Training of Word Embeddings Vector Completed#########"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XakWbPClBg_z"
      },
      "source": [
        "import numpy as np\n",
        "def getVectors(dataset):\n",
        "  singleDataItemEmbedding=np.zeros(embeddingsSize)\n",
        "  vectors=[]\n",
        "  for dataItem in dataset:\n",
        "    wordCount=0\n",
        "    for word in dataItem:\n",
        "      if word in model.wv.vocab:\n",
        "        singleDataItemEmbedding=singleDataItemEmbedding+model.wv[word]\n",
        "        wordCount=wordCount+1\n",
        "  \n",
        "    singleDataItemEmbedding=singleDataItemEmbedding/wordCount  \n",
        "    vectors.append(singleDataItemEmbedding)\n",
        "  return vectors\n",
        "\n",
        "trainReviewVectors=getVectors(trainReviews)\n",
        "testReviewVectors=getVectors(testReviews)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WhobrCpvDrUc"
      },
      "source": [
        "#Let's define a function that can display the accuracy, F1-score, label-wise precision, recall, etc. of each classifier\n",
        "\n",
        "from sklearn.metrics import accuracy_score\n",
        "#add path of google drive to environment variable to load python files from google drive\n",
        "import sys\n",
        "sys.path.insert(1, \"/content/drive/My Drive/nlp assignments/assignment 4\")\n",
        "from visualization import plot_confusion_matrix_from_data\n",
        "from sklearn.metrics import precision_recall_fscore_support as score\n",
        "from sklearn.metrics import f1_score\n",
        "\n",
        "def printResults(y_true, y_predicted):\n",
        "  print(\"Accuracy= \", accuracy_score(y_true, y_predicted))\n",
        "\n",
        "  columns=['false', 'true']\n",
        "  plot_confusion_matrix_from_data(y_true, y_predicted, columns)\n",
        "\n",
        "  precision, recall, fscore, support = score(y_true, y_predicted)\n",
        "\n",
        "  print('###########################################')\n",
        "  print('precision: {}'.format(precision))  \n",
        "  print('recall: {}'.format(recall))\n",
        "  print('fscore: {}'.format(fscore))\n",
        "  print('support: {}'.format(support))\n",
        "  print('###########################################3')\n",
        "\n",
        "  print('Macro F1 ',f1_score(y_true, y_predicted, average='macro'))\n",
        "\n",
        "  print('Micro F1 ', f1_score(y_true, y_predicted, average='micro'))\n",
        "\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x_nAxHEZCEtz",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "bd3c382d-4a45-43c5-f2fa-cc6c6207e287"
      },
      "source": [
        "#naive bayes\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "clfNB = MultinomialNB()\n",
        "\n",
        "scaler = MinMaxScaler()\n",
        "scaledTrainX= scaler.fit_transform(trainReviewVectors)\n",
        "scaledTestX = scaler.fit_transform(testReviewVectors)\n",
        "clfNB.fit(scaledTrainX, trainLabels)\n",
        "\n",
        "#test naive bayes accuracy\n",
        "testLabelsPredicted=list(clfNB.predict(scaledTestX))\n",
        "\n",
        "#print results\n",
        "print(\"####################RESULTS OF NAIVE BAYES CLASSIFIER##################\")\n",
        "printResults(testLabelsPredicted, testLabels)\n"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "####################RESULTS OF NAIVE BAYES CLASSIFIER##################\n",
            "Accuracy=  0.63264\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:165: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:166: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:168: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:169: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x648 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "###########################################\n",
            "precision: [0.55863402 0.70565798]\n",
            "recall: [0.6518797 0.6183844]\n",
            "fscore: [0.60166551 0.65914489]\n",
            "support: [2660 3590]\n",
            "###########################################3\n",
            "Macro F1  0.6304052015870478\n",
            "Micro F1  0.63264\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xW8P1L7eIKf3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "4ca1a009-2547-4d1a-e5c3-11fe7ea52d9d"
      },
      "source": [
        "#neural network\n",
        "from sklearn.neural_network import MLPClassifier\n",
        "\n",
        "clfMLP = MLPClassifier(hidden_layer_sizes=(10, 10, 10))\n",
        "clfMLP.fit(trainReviewVectors, trainLabels)\n",
        "  \n",
        "testLabelsPredicted=list(clfMLP.predict(testReviewVectors))\n",
        "\n",
        "#print results\n",
        "print(\"####################RESULTS OF NEURAL NETWORK CLASSIFIER##################\")\n",
        "printResults(testLabelsPredicted, testLabels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "####################RESULTS OF NEURAL NETWORK CLASSIFIER##################\n",
            "Accuracy=  0.81424\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:165: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:166: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:168: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:169: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x648 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "###########################################\n",
            "precision: [0.76739691 0.86045772]\n",
            "recall: [0.84438143 0.78944299]\n",
            "fscore: [0.80405063 0.82342205]\n",
            "support: [2821 3429]\n",
            "###########################################3\n",
            "Macro F1  0.8137363430716658\n",
            "Micro F1  0.81424\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uUJMx8OOZA66",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "0e16a5f7-a633-42cb-a617-7ded5ab5cd6f"
      },
      "source": [
        "# Random Forest Classifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "clfRF = RandomForestClassifier(n_estimators = 1000)\n",
        "# Train the model on training data\n",
        "clfRF.fit(trainReviewVectors, trainLabels);\n",
        "\n",
        "testLabelsPredicted=list(clfRF.predict(testReviewVectors))\n",
        "\n",
        "#print results\n",
        "print(\"####################RESULTS OF Random Forest CLASSIFIER##################\")\n",
        "printResults(testLabelsPredicted, testLabels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "####################RESULTS OF Random Forest CLASSIFIER##################\n",
            "Accuracy=  0.76688\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:165: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:166: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:168: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:169: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x648 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "###########################################\n",
            "precision: [0.75       0.78353465]\n",
            "recall: [0.77367896 0.76056773]\n",
            "fscore: [0.76165549 0.77188038]\n",
            "support: [3009 3241]\n",
            "###########################################3\n",
            "Macro F1  0.7667679351648027\n",
            "Micro F1  0.76688\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W68M8x0v_lDM",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "7bc2cc27-ab9f-43a7-bfc9-d52b804f4268"
      },
      "source": [
        "#KNN classifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "\n",
        "clfKNN = KNeighborsClassifier(n_neighbors=3)\n",
        "\n",
        "# Train the model on training data\n",
        "clfKNN.fit(trainReviewVectors, trainLabels);\n",
        "\n",
        "testLabelsPredicted=list(clfKNN.predict(testReviewVectors))\n",
        "\n",
        "#print results\n",
        "print(\"####################RESULTS OF KNN Classifier##################\")\n",
        "printResults(testLabelsPredicted, testLabels)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "####################RESULTS OF KNN Classifier##################\n",
            "Accuracy=  0.69248\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:165: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:166: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:168: MatplotlibDeprecationWarning: \n",
            "The tick1On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick1line.set_visible instead.\n",
            "  t.tick1On = False\n",
            "/content/drive/My Drive/nlp assignments/assignment 4/visualization.py:169: MatplotlibDeprecationWarning: \n",
            "The tick2On function was deprecated in Matplotlib 3.1 and will be removed in 3.3. Use Tick.tick2line.set_visible instead.\n",
            "  t.tick2On = False\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x648 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "###########################################\n",
            "precision: [0.71585052 0.66942149]\n",
            "recall: [0.68117719 0.70481928]\n",
            "fscore: [0.69808357 0.68666449]\n",
            "support: [3262 2988]\n",
            "###########################################3\n",
            "Macro F1  0.6923740309749964\n",
            "Micro F1  0.69248\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}