04_LSTM.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classification using LSTM\n",
    "\n",
    "## Getting started"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import f1_score, accuracy_score\n",
    "\n",
    "from src.utils.viz import *\n",
    "\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read & split train test\n",
    "filename = './data/clean_entire_corpus.csv'\n",
    "df = pd.read_csv(filename)\n",
    "df_train = df[df['source'] == 'train'].copy()\n",
    "df_test = df[df['source'] == 'test'].copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Tokenize data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Encoding\n",
    "def get_vocabulary(iter_sents):\n",
    "    vocab = set()\n",
    "    max_length = 0\n",
    "    for text in iter_sents:\n",
    "        tokens = text.split(' ')\n",
    "        if len(tokens) > max_length:\n",
    "            max_length = len(tokens)\n",
    "        vocab.update(tokens)\n",
    "    return vocab, max_length\n",
    "\n",
    "vocab, max_length = get_vocabulary(df_train['clean_text'])\n",
    "\n",
    "tokenizer = tf.keras.preprocessing.text.Tokenizer()\n",
    "tokenizer.fit_on_texts(df_train['clean_text'])\n",
    "sequences_train = tokenizer.texts_to_sequences(df_train['clean_text'])\n",
    "sequences_test = tokenizer.texts_to_sequences(df_test['clean_text'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Padding texts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Padding\n",
    "padded_seqs_train = tf.keras.preprocessing.sequence.pad_sequences(sequences_train, maxlen=max_length,\n",
    "                                                            padding='post', truncating='post')\n",
    "padded_seqs_test = tf.keras.preprocessing.sequence.pad_sequences(sequences_test, maxlen=max_length,\n",
    "                                                            padding='post', truncating='post')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Convert padded texts to tf.data.Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data sets\n",
    "X_train = padded_seqs_train\n",
    "X_test = padded_seqs_test\n",
    "y_train = df_train['target'].values\n",
    "y_test = df_test['target'].values\n",
    "\n",
    "train_ds = tf.data.Dataset.from_tensor_slices((X_train, y_train)).shuffle(1000).batch(100)\n",
    "test_ds = tf.data.Dataset.from_tensor_slices((X_test, y_test)).batch(100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Import GloVe embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Reading lines\n",
      "Extract embeddings\n",
      "Create embeddings matrix\n"
     ]
    }
   ],
   "source": [
    "# Build an index for embeddings\n",
    "embeddings_index = {}\n",
    "EMBEDDINGS_PATH = '/datascience/embeddings'\n",
    "EMBEDDINGS_LENGTH = 50\n",
    "print('Reading lines')\n",
    "with open(os.path.join(EMBEDDINGS_PATH, 'glove.6B.50d.txt'), encoding='utf-8') as f:\n",
    "    lines = f.readlines()\n",
    "\n",
    "print('Extract embeddings')\n",
    "for line in lines:\n",
    "    values = line.split()\n",
    "    word = values[0]\n",
    "    coefs = np.asarray(values[1:], dtype='float32')\n",
    "    embeddings_index[word] = coefs\n",
    "\n",
    "print('Create embeddings matrix')\n",
    "word_index = tokenizer.word_index\n",
    "embedding_matrix = np.zeros((len(word_index) + 1, EMBEDDINGS_LENGTH))\n",
    "for word, i in word_index.items():\n",
    "    embedding_vector = embeddings_index.get(word)\n",
    "    if embedding_vector is not None:\n",
    "        # words not found in embedding index will be all-zeros.\n",
    "        embedding_matrix[i] = embedding_vector\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "69/69 [==============================] - 4s 58ms/step - loss: 0.5407 - acc: 0.7364 - val_loss: 0.4890 - val_acc: 0.7882\n",
      "Epoch 2/10\n",
      "69/69 [==============================] - 4s 51ms/step - loss: 0.4546 - acc: 0.8031 - val_loss: 0.4629 - val_acc: 0.7882\n",
      "Epoch 3/10\n",
      "69/69 [==============================] - 4s 53ms/step - loss: 0.4386 - acc: 0.8045 - val_loss: 0.4603 - val_acc: 0.7922\n",
      "Epoch 4/10\n",
      "69/69 [==============================] - 4s 51ms/step - loss: 0.4317 - acc: 0.8095 - val_loss: 0.4549 - val_acc: 0.7925\n",
      "Epoch 5/10\n",
      "69/69 [==============================] - 4s 52ms/step - loss: 0.4240 - acc: 0.8159 - val_loss: 0.4809 - val_acc: 0.7846\n",
      "Epoch 6/10\n",
      "69/69 [==============================] - 4s 51ms/step - loss: 0.4201 - acc: 0.8158 - val_loss: 0.4613 - val_acc: 0.7959\n",
      "Epoch 7/10\n",
      "69/69 [==============================] - 4s 53ms/step - loss: 0.4227 - acc: 0.8181 - val_loss: 0.4864 - val_acc: 0.7980\n",
      "Epoch 8/10\n",
      "69/69 [==============================] - 4s 51ms/step - loss: 0.4110 - acc: 0.8235 - val_loss: 0.4509 - val_acc: 0.7974\n",
      "Epoch 9/10\n",
      "69/69 [==============================] - 4s 52ms/step - loss: 0.4011 - acc: 0.8259 - val_loss: 0.4942 - val_acc: 0.7904\n",
      "Epoch 10/10\n",
      "69/69 [==============================] - 4s 51ms/step - loss: 0.3920 - acc: 0.8326 - val_loss: 0.4594 - val_acc: 0.7996\n"
     ]
    }
   ],
   "source": [
    "# Model\n",
    "model = tf.keras.Sequential()\n",
    "model.add(p := tf.keras.layers.Embedding(input_dim=len(word_index)+1, output_dim=EMBEDDINGS_LENGTH,\n",
    "                                         input_length=max_length, weights=[embedding_matrix], trainable=False))\n",
    "model.add(a := tf.keras.layers.LSTM(128))\n",
    "model.add(b := tf.keras.layers.Dropout(0.2))\n",
    "model.add(c := tf.keras.layers.Flatten())\n",
    "model.add(d := tf.keras.layers.Dense(64, activation='relu'))\n",
    "model.add(h := tf.keras.layers.Dense(1, activation='sigmoid'))\n",
    "\n",
    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.001)\n",
    "model.compile(optimizer=optimizer, loss='binary_crossentropy', metrics=['acc'])\n",
    "\n",
    "EPOCHS = 10\n",
    "history = model.fit(train_ds, epochs=EPOCHS, validation_data=test_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1dd0f0db940>"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "epochs = range(1, EPOCHS+1)\n",
    "fig, axs = plt.subplots(1, 2, figsize=(25, 8))\n",
    "\n",
    "axs[0].plot(epochs, history.history['acc'], color='blue', label='train')\n",
    "axs[0].plot(epochs, history.history['val_acc'], color='orange', label='test')\n",
    "axs[0].set_xticks(range(1, EPOCHS+1, 5))\n",
    "axs[0].set_xlabel('Epochs')\n",
    "axs[0].set_ylabel('Accuracy')\n",
    "\n",
    "axs[1].plot(epochs, history.history['loss'], color='blue', label='train')\n",
    "axs[1].plot(epochs, history.history['val_loss'], color='orange', label='test')\n",
    "axs[1].set_xticks(range(1, EPOCHS+1, 5))\n",
    "axs[1].set_xlabel('Epochs')\n",
    "axs[1].set_ylabel('Loss')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* The best score reaches 0.80 of accuracy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model's performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(<Figure size 432x432 with 2 Axes>,\n",
       " <matplotlib.axes._subplots.AxesSubplot at 0x1dd2a1e00a0>)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred = model.predict(X_test).reshape(1, -1)[0]\n",
    "y_pred = np.around(y_pred).astype(int)\n",
    "\n",
    "plot_confusion_matrix(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x1dd164ef340>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA\n",
    "\n",
    "embed = lambda x: d(c(b(a(p(x)))))\n",
    "embedded_text = embed(X_test)\n",
    "\n",
    "twodim = PCA(2).fit_transform(embedded_text)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(10, 10))\n",
    "ax.scatter(twodim[:,0], twodim[:,1], c=y_test, alpha=.6)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model's mistakes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "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>id</th>\n",
       "      <th>keyword</th>\n",
       "      <th>text</th>\n",
       "      <th>target</th>\n",
       "      <th>source</th>\n",
       "      <th>clean_text</th>\n",
       "      <th>pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6857</th>\n",
       "      <td>51</td>\n",
       "      <td>ablaze</td>\n",
       "      <td>#PreviouslyOnDoyinTv: Toke MakinwaÛªs marriag...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>previously on doyin tv toke makinwas marriage...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6861</th>\n",
       "      <td>70</td>\n",
       "      <td>ablaze</td>\n",
       "      <td>Burning Man Ablaze! by Turban Diva http://t.co...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>burning man ablaze by turban diva via</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6867</th>\n",
       "      <td>90</td>\n",
       "      <td>ablaze</td>\n",
       "      <td>'Burning Rahm': Let's hope City Hall builds a ...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>burning rahm let is hope city hall builds a g...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6872</th>\n",
       "      <td>106</td>\n",
       "      <td>accident</td>\n",
       "      <td>#3: Car Recorder ZeroEdgeå¨ Dual-lens Car Came...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>car recorder zeroedge dual lens car camera ve...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6873</th>\n",
       "      <td>108</td>\n",
       "      <td>accident</td>\n",
       "      <td>Coincidence Or #Curse? Still #Unresolved Secre...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>coincidence or curse still unresolved secrets ...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10084</th>\n",
       "      <td>10797</td>\n",
       "      <td>wrecked</td>\n",
       "      <td>@Nathan26_RFC thought you said Saturday night ...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>thought you said saturday night there and nea...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10086</th>\n",
       "      <td>10804</td>\n",
       "      <td>wrecked</td>\n",
       "      <td>@yakubOObs think he deactivated because his no...</td>\n",
       "      <td>1</td>\n",
       "      <td>test</td>\n",
       "      <td>think he deactivated because his notification...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10087</th>\n",
       "      <td>10806</td>\n",
       "      <td>wrecked</td>\n",
       "      <td>RT CNBC '3 words from Disney CEO Bob Iger wrec...</td>\n",
       "      <td>1</td>\n",
       "      <td>test</td>\n",
       "      <td>rt cnbc words from disney ceo bob iger wrecked...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10096</th>\n",
       "      <td>10857</td>\n",
       "      <td>NaN</td>\n",
       "      <td>To conference attendees! The blue line from th...</td>\n",
       "      <td>1</td>\n",
       "      <td>test</td>\n",
       "      <td>to conference attendees the blue line from the...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10098</th>\n",
       "      <td>10861</td>\n",
       "      <td>NaN</td>\n",
       "      <td>EARTHQUAKE SAFETY LOS ANGELES ÛÒ SAFETY FASTE...</td>\n",
       "      <td>0</td>\n",
       "      <td>test</td>\n",
       "      <td>earthquake safety los angeles safety fasteners...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>654 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          id   keyword                                               text  \\\n",
       "6857      51    ablaze  #PreviouslyOnDoyinTv: Toke MakinwaÛªs marriag...   \n",
       "6861      70    ablaze  Burning Man Ablaze! by Turban Diva http://t.co...   \n",
       "6867      90    ablaze  'Burning Rahm': Let's hope City Hall builds a ...   \n",
       "6872     106  accident  #3: Car Recorder ZeroEdgeå¨ Dual-lens Car Came...   \n",
       "6873     108  accident  Coincidence Or #Curse? Still #Unresolved Secre...   \n",
       "...      ...       ...                                                ...   \n",
       "10084  10797   wrecked  @Nathan26_RFC thought you said Saturday night ...   \n",
       "10086  10804   wrecked  @yakubOObs think he deactivated because his no...   \n",
       "10087  10806   wrecked  RT CNBC '3 words from Disney CEO Bob Iger wrec...   \n",
       "10096  10857       NaN  To conference attendees! The blue line from th...   \n",
       "10098  10861       NaN  EARTHQUAKE SAFETY LOS ANGELES ÛÒ SAFETY FASTE...   \n",
       "\n",
       "       target source                                         clean_text  pred  \n",
       "6857        0   test   previously on doyin tv toke makinwas marriage...     1  \n",
       "6861        0   test             burning man ablaze by turban diva via      1  \n",
       "6867        0   test   burning rahm let is hope city hall builds a g...     1  \n",
       "6872        0   test   car recorder zeroedge dual lens car camera ve...     1  \n",
       "6873        0   test  coincidence or curse still unresolved secrets ...     1  \n",
       "...       ...    ...                                                ...   ...  \n",
       "10084       0   test   thought you said saturday night there and nea...     1  \n",
       "10086       1   test   think he deactivated because his notification...     0  \n",
       "10087       1   test  rt cnbc words from disney ceo bob iger wrecked...     0  \n",
       "10096       1   test  to conference attendees the blue line from the...     0  \n",
       "10098       0   test  earthquake safety los angeles safety fasteners...     1  \n",
       "\n",
       "[654 rows x 7 columns]"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test['pred'] = y_pred\n",
    "model_errors = df_test[df_test['pred'] != df_test['target']]\n",
    "model_errors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1dd2a1d2c40>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "top_unigrams = get_top_grams(model_errors['clean_text'], 20)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(18, 8))\n",
    "\n",
    "df1 = pd.DataFrame(top_unigrams, columns = ['text' , 'count'])\n",
    "df1['count'] = df1['count'].astype(float)\n",
    "df1.plot(x='text', y='count', kind='bar', ax=ax, title='Unigram frequency')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1dd0bdc34c0>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "top_bigrams = get_top_grams(model_errors['clean_text'], 20, gram=2)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(18, 8))\n",
    "\n",
    "df1 = pd.DataFrame(top_bigrams, columns = ['text' , 'count'])\n",
    "df1['count'] = df1['count'].astype(float)\n",
    "df1.plot(x='text', y='count', kind='bar', ax=ax, title='Bigram frequency')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1dd16379040>"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1296x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "top_bigrams = get_top_grams(model_errors['clean_text'], 20, gram=3)\n",
    "fig, ax = plt.subplots(1, 1, figsize=(18, 8))\n",
    "\n",
    "df1 = pd.DataFrame(top_bigrams, columns = ['text' , 'count'])\n",
    "df1['count'] = df1['count'].astype(float)\n",
    "df1.plot(x='text', y='count', kind='bar', ax=ax, title='Trigram frequency')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "nlp",
   "language": "python",
   "name": "nlp"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}