CNN+LSTM_Sentiment_Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "from keras.datasets import imdb"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded dataset with 25000 training samples, 25000 test samples\n"
     ]
    }
   ],
   "source": [
    "vocabulary_size = 5000\n",
    "(X_train, y_train), (X_test, y_test) = imdb.load_data(num_words = vocabulary_size)\n",
    "print('Loaded dataset with {} training samples, {} test samples'.format(len(X_train), len(X_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---review---\n",
      "[1, 2, 365, 1234, 5, 1156, 354, 11, 14, 2, 2, 7, 1016, 2, 2, 356, 44, 4, 1349, 500, 746, 5, 200, 4, 4132, 11, 2, 2, 1117, 1831, 2, 5, 4831, 26, 6, 2, 4183, 17, 369, 37, 215, 1345, 143, 2, 5, 1838, 8, 1974, 15, 36, 119, 257, 85, 52, 486, 9, 6, 2, 2, 63, 271, 6, 196, 96, 949, 4121, 4, 2, 7, 4, 2212, 2436, 819, 63, 47, 77, 2, 180, 6, 227, 11, 94, 2494, 2, 13, 423, 4, 168, 7, 4, 22, 5, 89, 665, 71, 270, 56, 5, 13, 197, 12, 161, 2, 99, 76, 23, 2, 7, 419, 665, 40, 91, 85, 108, 7, 4, 2084, 5, 4773, 81, 55, 52, 1901]\n",
      "---label---\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "print('---review---')\n",
    "print(X_train[6])\n",
    "print('---label---')\n",
    "print(y_train[6])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "---review with words---\n",
      "['the', 'and', 'full', 'involving', 'to', 'impressive', 'boring', 'this', 'as', 'and', 'and', 'br', 'villain', 'and', 'and', 'need', 'has', 'of', 'costumes', 'b', 'message', 'to', 'may', 'of', 'props', 'this', 'and', 'and', 'concept', 'issue', 'and', 'to', \"god's\", 'he', 'is', 'and', 'unfolds', 'movie', 'women', 'like', \"isn't\", 'surely', \"i'm\", 'and', 'to', 'toward', 'in', \"here's\", 'for', 'from', 'did', 'having', 'because', 'very', 'quality', 'it', 'is', 'and', 'and', 'really', 'book', 'is', 'both', 'too', 'worked', 'carl', 'of', 'and', 'br', 'of', 'reviewer', 'closer', 'figure', 'really', 'there', 'will', 'and', 'things', 'is', 'far', 'this', 'make', 'mistakes', 'and', 'was', \"couldn't\", 'of', 'few', 'br', 'of', 'you', 'to', \"don't\", 'female', 'than', 'place', 'she', 'to', 'was', 'between', 'that', 'nothing', 'and', 'movies', 'get', 'are', 'and', 'br', 'yes', 'female', 'just', 'its', 'because', 'many', 'br', 'of', 'overly', 'to', 'descent', 'people', 'time', 'very', 'bland']\n",
      "---label---\n",
      "1\n"
     ]
    }
   ],
   "source": [
    "word2id = imdb.get_word_index()\n",
    "id2word = {i: word for word, i in word2id.items()}\n",
    "print('---review with words---')\n",
    "print([id2word.get(i) for i in X_train[6]])\n",
    "print('---label---')\n",
    "print(y_train[6])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Maximum review length: 2697\n"
     ]
    }
   ],
   "source": [
    "print('Maximum review length: {}'.format(\n",
    "len(max((X_train + X_test), key=len))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum review length: 14\n"
     ]
    }
   ],
   "source": [
    "print('Minimum review length: {}'.format(\n",
    "len(min((X_test + X_test), key=len))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.preprocessing import sequence\n",
    "max_words = 200\n",
    "X_train = sequence.pad_sequences(X_train, maxlen=max_words)\n",
    "X_test = sequence.pad_sequences(X_test, maxlen=max_words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   5,   25,  100,   43,  838,  112,   50,  670,    2,    9,   35,\n",
       "        480,  284,    5,  150,    4,  172,  112,  167,    2,  336,  385,\n",
       "         39,    4,  172, 4536, 1111,   17,  546,   38,   13,  447,    4,\n",
       "        192,   50,   16,    6,  147, 2025,   19,   14,   22,    4, 1920,\n",
       "       4613,  469,    4,   22,   71,   87,   12,   16,   43,  530,   38,\n",
       "         76,   15,   13, 1247,    4,   22,   17,  515,   17,   12,   16,\n",
       "        626,   18,    2,    5,   62,  386,   12,    8,  316,    8,  106,\n",
       "          5,    4, 2223,    2,   16,  480,   66, 3785,   33,    4,  130,\n",
       "         12,   16,   38,  619,    5,   25,  124,   51,   36,  135,   48,\n",
       "         25, 1415,   33,    6,   22,   12,  215,   28,   77,   52,    5,\n",
       "         14,  407,   16,   82,    2,    8,    4,  107,  117,    2,   15,\n",
       "        256,    4,    2,    7, 3766,    5,  723,   36,   71,   43,  530,\n",
       "        476,   26,  400,  317,   46,    7,    4,    2, 1029,   13,  104,\n",
       "         88,    4,  381,   15,  297,   98,   32, 2071,   56,   26,  141,\n",
       "          6,  194,    2,   18,    4,  226,   22,   21,  134,  476,   26,\n",
       "        480,    5,  144,   30,    2,   18,   51,   36,   28,  224,   92,\n",
       "         25,  104,    4,  226,   65,   16,   38, 1334,   88,   12,   16,\n",
       "        283,    5,   16, 4472,  113,  103,   32,   15,   16,    2,   19,\n",
       "        178,   32])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from keras import Sequential\n",
    "# from keras.layers import Embedding, LSTM, Dense, Dropout, Conv1D, MaxPooling1D\n",
    "# embedding_size=200\n",
    "# model=Sequential()\n",
    "# model.add(Embedding(vocabulary_size, embedding_size, input_length=max_words))\n",
    "# model.add(Conv1D(32, kernel_size=5,\n",
    "#                  activation='relu'))\n",
    "# model.add(MaxPooling1D(pool_size=2))\n",
    "# model.add(Dropout(0.5))\n",
    "# model.add(LSTM(100))\n",
    "# model.add(Dropout(0.5))\n",
    "# model.add(Dense(1, activation='sigmoid'))\n",
    "# print(model.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers import Embedding\n",
    "from keras.models import Sequential, Model\n",
    "from keras.layers import Dense, Activation\n",
    "from keras.layers import Flatten, Conv1D, SpatialDropout1D, MaxPooling1D,AveragePooling1D, Bidirectional, merge, concatenate, Input, Dropout, LSTM\n",
    "\n",
    "def model(embedding_size=200, max_words=200, y_dim=1, num_filters=200, filter_sizes=[3,4,5], pool_padding='valid', dropout=0.5):\n",
    "    embed_input = Input(shape=(max_words,))\n",
    "    x = Embedding(vocabulary_size, embedding_size, input_length=max_words)(embed_input)\n",
    "    pooled_outputs = []\n",
    "    for i in range(len(filter_sizes)):\n",
    "        conv = Conv1D(num_filters, kernel_size=filter_sizes[i], padding='valid', activation='relu')(x)\n",
    "        conv = MaxPooling1D(pool_size=max_words-filter_sizes[i]+1)(conv)           \n",
    "        pooled_outputs.append(conv)\n",
    "    merge = concatenate(pooled_outputs)\n",
    "    \n",
    "    x = Dense(30, activation='relu')(merge)\n",
    "    x = Dropout(dropout)(x)\n",
    "    x = Bidirectional(LSTM(100, return_sequences=True, dropout=0.5, recurrent_dropout=0.1))(x)\n",
    "    x = Dense(30, activation='relu')(x)\n",
    "    x = Dropout(dropout)(x)\n",
    "    x = Flatten()(x)\n",
    "    x = Dense(y_dim, activation='sigmoid')(x)\n",
    "\n",
    "    model = Model(inputs=embed_input,outputs=x)\n",
    "\n",
    "#   model.compile(optimizer='adam',loss = 'categorical_crossentropy', metrics = ['acc'])\n",
    "    print(model.summary())\n",
    "    \n",
    "    from keras.utils import plot_model\n",
    "    plot_model(model, to_file='shared_input_layer.png')\n",
    "    \n",
    "    return model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 200)          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "embedding_1 (Embedding)         (None, 200, 200)     1000000     input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_1 (Conv1D)               (None, 198, 200)     120200      embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_2 (Conv1D)               (None, 197, 200)     160200      embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "conv1d_3 (Conv1D)               (None, 196, 200)     200200      embedding_1[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_1 (MaxPooling1D)  (None, 1, 200)       0           conv1d_1[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_2 (MaxPooling1D)  (None, 1, 200)       0           conv1d_2[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "max_pooling1d_3 (MaxPooling1D)  (None, 1, 200)       0           conv1d_3[0][0]                   \n",
      "__________________________________________________________________________________________________\n",
      "concatenate_1 (Concatenate)     (None, 1, 600)       0           max_pooling1d_1[0][0]            \n",
      "                                                                 max_pooling1d_2[0][0]            \n",
      "                                                                 max_pooling1d_3[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 1, 30)        18030       concatenate_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "dropout_1 (Dropout)             (None, 1, 30)        0           dense_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "bidirectional_1 (Bidirectional) (None, 1, 200)       104800      dropout_1[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 1, 30)        6030        bidirectional_1[0][0]            \n",
      "__________________________________________________________________________________________________\n",
      "dropout_2 (Dropout)             (None, 1, 30)        0           dense_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "flatten_1 (Flatten)             (None, 30)           0           dropout_2[0][0]                  \n",
      "__________________________________________________________________________________________________\n",
      "dense_3 (Dense)                 (None, 1)            31          flatten_1[0][0]                  \n",
      "==================================================================================================\n",
      "Total params: 1,609,491\n",
      "Trainable params: 1,609,491\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = model(embedding_size=200, max_words=200,y_dim=1,filter_sizes = [3,4,5],dropout=0.5)\n",
    "from IPython.display import Image\n",
    "Image(filename='shared_input_layer.png') "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='binary_crossentropy', \n",
    "             optimizer='adam', \n",
    "             metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 24936 samples, validate on 64 samples\n",
      "Epoch 1/3\n",
      "24936/24936 [==============================] - 32s 1ms/step - loss: 0.5412 - acc: 0.7042 - val_loss: 0.2903 - val_acc: 0.8594\n",
      "Epoch 2/3\n",
      "24936/24936 [==============================] - 27s 1ms/step - loss: 0.3498 - acc: 0.8521 - val_loss: 0.2512 - val_acc: 0.9219\n",
      "Epoch 3/3\n",
      "24936/24936 [==============================] - 28s 1ms/step - loss: 0.2617 - acc: 0.9012 - val_loss: 0.2756 - val_acc: 0.8906\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1227bde66d8>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "batch_size = 64\n",
    "num_epochs = 3\n",
    "X_valid, y_valid = X_train[:batch_size], y_train[:batch_size]\n",
    "X_train2, y_train2 = X_train[batch_size:], y_train[batch_size:]\n",
    "model.fit(X_train2, y_train2, validation_data=(X_valid, y_valid), batch_size=batch_size, epochs=num_epochs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test accuracy: 0.88544\n"
     ]
    }
   ],
   "source": [
    "scores = model.evaluate(X_test, y_test, verbose=0)\n",
    "print('Test accuracy:', scores[1])"
   ]
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}