Amazon_Reviews_for_Sentiment_Analysis.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Amazon Reviews for Sentiment Analysis.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "B1_Lis8_furD",
        "riCk874siEqN"
      ]
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "#Import Libraries "
      ],
      "metadata": {
        "id": "3WmbQC2ngiQU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.metrics import accuracy_score,confusion_matrix,classification_report\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from keras import models, layers, optimizers\n",
        "from keras.preprocessing.text import Tokenizer, text_to_word_sequence\n",
        "from keras.preprocessing.sequence import pad_sequences\n",
        "import bz2\n",
        "from sklearn.metrics import f1_score, roc_auc_score, accuracy_score\n",
        "import re\n",
        "import nltk\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')\n",
        "from nltk.stem import PorterStemmer\n",
        "from nltk.tokenize import word_tokenize\n",
        "from nltk.corpus import stopwords\n",
        "import pickle \n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "from tensorflow.keras.layers import Embedding,SimpleRNN,Dense,LSTM,Dropout,Conv1D,MaxPooling1D,Flatten,BatchNormalization,Bidirectional,Lambda,GRU\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.keras.models import Sequential\n",
        "from sklearn.model_selection import train_test_split"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9aImzJuogohc",
        "outputId": "3fb01683-3876-43ca-ac93-08979e702fbf"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n",
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Download The Dataset"
      ],
      "metadata": {
        "id": "B1_Lis8_furD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!wget 'https://storage.googleapis.com/kaggle-data-sets/1305/800230/bundle/archive.zip?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com%40kaggle-161607.iam.gserviceaccount.com%2F20220504%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20220504T143549Z&X-Goog-Expires=259199&X-Goog-SignedHeaders=host&X-Goog-Signature=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'"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4B2p89iAfx_Z",
        "outputId": "cc5cd7ed-006a-4bd8-880b-2a63add931fe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The name is too long, 767 chars total.\n",
            "Trying to shorten...\n",
            "New name is archive.zip?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com@kaggle-161607.iam.gserviceaccount.com%2F20220504%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20220504T143549Z&X-Goog-Expires=259199&X-Goog-SignedHeaders=hos.\n",
            "--2022-05-04 16:45:18--  https://storage.googleapis.com/kaggle-data-sets/1305/800230/bundle/archive.zip?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com%40kaggle-161607.iam.gserviceaccount.com%2F20220504%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20220504T143549Z&X-Goog-Expires=259199&X-Goog-SignedHeaders=host&X-Goog-Signature=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\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 74.125.141.128, 173.194.211.128, 74.125.134.128, ...\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|74.125.141.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 517080965 (493M) [application/zip]\n",
            "Saving to: ‘archive.zip?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com@kaggle-161607.iam.gserviceaccount.com%2F20220504%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20220504T143549Z&X-Goog-Expires=259199&X-Goog-SignedHeaders=hos’\n",
            "\n",
            "archive.zip?X-Goog- 100%[===================>] 493.13M   100MB/s    in 4.9s    \n",
            "\n",
            "2022-05-04 16:45:23 (100 MB/s) - ‘archive.zip?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=gcp-kaggle-com@kaggle-161607.iam.gserviceaccount.com%2F20220504%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20220504T143549Z&X-Goog-Expires=259199&X-Goog-SignedHeaders=hos’ saved [517080965/517080965]\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!unzip /content/archive.zip"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qZ5Nx9nTfzUE",
        "outputId": "bc3a7da5-7dec-488f-cdce-e30093b29006"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Archive:  /content/archive.zip\n",
            "  inflating: test.ft.txt.bz2         \n",
            "  inflating: train.ft.txt.bz2        \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def get_labels_and_texts(file):\n",
        "    labels = []\n",
        "    texts = []\n",
        "    for line in bz2.BZ2File(file):\n",
        "        x = line.decode(\"utf-8\")\n",
        "        labels.append(int(x[9]) - 1)\n",
        "        texts.append(x[10:].strip())\n",
        "    return np.array(labels), texts\n",
        "train_labels, train_texts = get_labels_and_texts('../content/train.ft.txt.bz2')\n",
        "test_labels, test_texts = get_labels_and_texts('../content/test.ft.txt.bz2')"
      ],
      "metadata": {
        "id": "d_v98QnjgIB1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_labels.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UfddFlovhiBU",
        "outputId": "830d7d1a-af30-4024-8487-15fe358dab38"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(3600000,)"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_texts)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "F4JaJ4rihjMk",
        "outputId": "6eb8e26d-dc15-42dc-8352-c162e53671f1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3600000"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#PreProcessing"
      ],
      "metadata": {
        "id": "riCk874siEqN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "stop_words = set(stopwords.words('english'))\n",
        "ps = PorterStemmer()"
      ],
      "metadata": {
        "id": "JVefO3HoiB-U"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def filtered_text(texts):\n",
        "  texts_filtered = []\n",
        "  for train_text in (texts):\n",
        "    word_tokens = word_tokenize(train_text)\n",
        "    filtered_sentence = [ps.stem(w) for w in word_tokens if not w.lower() in stop_words]\n",
        "    texts_filtered.append(' '.join(filtered_sentence))\n",
        "  return texts_filtered"
      ],
      "metadata": {
        "id": "4XYOnUZQtUft"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_texts_filtered = filtered_text(train_texts)\n",
        "test_texts_filtered = filtered_text(test_texts)"
      ],
      "metadata": {
        "id": "YdbwOznci2cN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "max_word_within_sentence = 250"
      ],
      "metadata": {
        "id": "5gJ1gS8i_0MA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vocab_size = 10000\n",
        "embedding_dim = 100\n",
        "max_length = 300\n",
        "trunc_type='post'\n",
        "oov_tok = \"<OOV>\"\n",
        "\n",
        "tokenizer = Tokenizer(num_words = vocab_size, oov_token=oov_tok)\n",
        "tokenizer.fit_on_texts(train_texts_filtered)\n",
        "word_index = tokenizer.word_index\n",
        "\n",
        "train_sequences = tokenizer.texts_to_sequences(train_texts_filtered)\n",
        "X_train = pad_sequences(train_sequences,maxlen=max_length, truncating=trunc_type)\n",
        "\n",
        "test_sequences = tokenizer.texts_to_sequences(test_texts_filtered)\n",
        "X_test = pad_sequences(test_sequences,maxlen=max_length, truncating=trunc_type)"
      ],
      "metadata": {
        "id": "N_qMDVI2Majf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "path = '/content/drive/MyDrive/For Github/Amazon Reviews/'\n",
        "with open(path + 'X_train.pickle', 'wb') as handle:\n",
        "     pickle.dump(X_train,handle,protocol=pickle.HIGHEST_PROTOCOL)\n",
        "with open(path + 'X_test.pickle', 'wb') as handle:\n",
        "     pickle.dump(X_test,handle,protocol=pickle.HIGHEST_PROTOCOL)"
      ],
      "metadata": {
        "id": "88so2qqIhuQy"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#Define The Model"
      ],
      "metadata": {
        "id": "KmWAqB7Zgb4Z"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "I save thte X_train, X_test, y_train, y_test in my drive to use them directly "
      ],
      "metadata": {
        "id": "9bYNH8KydnpJ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "path = '/content/drive/MyDrive/For Github/Amazon Reviews/'"
      ],
      "metadata": {
        "id": "nwn87ESih1Rk"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "with open(path + 'X_train.pickle', 'rb') as handle:\n",
        "     X_train = pickle.load(handle)\n",
        "with open(path + 'X_test.pickle', 'rb') as handle:\n",
        "     X_test = pickle.load(handle)"
      ],
      "metadata": {
        "id": "SwBNz6YzgQSK"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "with open(path + 'train_labels.pickle', 'rb') as handle:\n",
        "    train_labels = pickle.load(handle)\n",
        "with open(path + 'test_labels.pickle', 'rb') as handle:\n",
        "    test_labels = pickle.load(handle)"
      ],
      "metadata": {
        "id": "UoF1HQk4hUBB"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class myCallback(tf.keras.callbacks.Callback):\n",
        "  def on_epoch_end(self, epoch, logs={}):\n",
        "    if(logs.get('accuracy')>=0.95):\n",
        "      print(\"\\nReached 95.0% accuracy so cancelling training!\")\n",
        "      self.model.stop_training = True\n",
        "callbacks = myCallback()"
      ],
      "metadata": {
        "id": "V8bjwdW2diuG"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vocab_size = 10000\n",
        "embedding_dim = 100\n",
        "max_length = 300"
      ],
      "metadata": {
        "id": "3Lb1ndvZhKuJ"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model=Sequential()\n",
        "model.add(Embedding(vocab_size,embedding_dim,input_length=max_length))\n",
        "model.add(BatchNormalization())\n",
        "model.add(LSTM(10))\n",
        "model.add(Dense(1,activation='sigmoid'))\n",
        "model.compile(loss='binary_crossentropy',optimizer=\"adam\",metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "pbYwmYVQcEoK"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "h = model.fit(X_train,train_labels,batch_size=1024,epochs=5,validation_data=(X_test,test_labels))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ir__HDzehOzx",
        "outputId": "6c4d9502-7162-417f-ab66-a5cd11084b62"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "3516/3516 [==============================] - 399s 113ms/step - loss: 0.2388 - accuracy: 0.9014 - val_loss: 0.2152 - val_accuracy: 0.9130\n",
            "Epoch 2/5\n",
            "3516/3516 [==============================] - 393s 112ms/step - loss: 0.2049 - accuracy: 0.9177 - val_loss: 0.2051 - val_accuracy: 0.9177\n",
            "Epoch 3/5\n",
            "3516/3516 [==============================] - 393s 112ms/step - loss: 0.1942 - accuracy: 0.9226 - val_loss: 0.2017 - val_accuracy: 0.9194\n",
            "Epoch 4/5\n",
            "3516/3516 [==============================] - 394s 112ms/step - loss: 0.1870 - accuracy: 0.9259 - val_loss: 0.1985 - val_accuracy: 0.9211\n",
            "Epoch 5/5\n",
            "3516/3516 [==============================] - 393s 112ms/step - loss: 0.1815 - accuracy: 0.9283 - val_loss: 0.1965 - val_accuracy: 0.9219\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y_pred = model.predict(X_test)\n",
        "y_pred = y_pred.round()"
      ],
      "metadata": {
        "id": "HeAMAOECHizP"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(classification_report(test_labels, y_pred))\n",
        "sns.heatmap(confusion_matrix(test_labels,y_pred), annot=True,fmt='g')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 439
        },
        "id": "mCM6bn79cF7x",
        "outputId": "1459ad85-43d0-46b8-eb3a-64f5aa13b5c6"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.92      0.92      0.92    200000\n",
            "           1       0.92      0.92      0.92    200000\n",
            "\n",
            "    accuracy                           0.92    400000\n",
            "   macro avg       0.92      0.92      0.92    400000\n",
            "weighted avg       0.92      0.92      0.92    400000\n",
            "\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7faab1a86c90>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(h.history['accuracy'])\n",
        "plt.plot(h.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(h.history['loss'])\n",
        "plt.plot(h.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "bivviGzms44p",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "outputId": "f2a82803-8e1b-4d35-b292-9924d4fcffc6"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "ETxzb1jIhSEa"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}