NeuralCF.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "# Training samples path, change to your local path\n",
    "training_samples_file_path = tf.keras.utils.get_file(\"trainingSamples.csv\", \"file:/Users/linchenxiao/Python_Code/Spark/SparrowRecsys/sampledata/trainingSamples.csv\")\n",
    "# Test samples path, change to your local path\n",
    "test_samples_file_path = tf.keras.utils.get_file(\"testSamples.csv\", \"file:/Users/linchenxiao/Python_Code/Spark/SparrowRecsys/sampledata/testSamples.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "7403/7403 [==============================] - 19s 2ms/step - loss: 0.6541 - accuracy: 0.6014 - auc_4: 0.6236 - auc_5: 0.6798\n",
      "Epoch 2/5\n",
      "7403/7403 [==============================] - 15s 2ms/step - loss: 0.5781 - accuracy: 0.6967 - auc_4: 0.7586 - auc_5: 0.7888\n",
      "Epoch 3/5\n",
      "7403/7403 [==============================] - 17s 2ms/step - loss: 0.5247 - accuracy: 0.7386 - auc_4: 0.8111 - auc_5: 0.8369\n",
      "Epoch 4/5\n",
      "7403/7403 [==============================] - 16s 2ms/step - loss: 0.4751 - accuracy: 0.7734 - auc_4: 0.8506 - auc_5: 0.8739\n",
      "Epoch 5/5\n",
      "7403/7403 [==============================] - 16s 2ms/step - loss: 0.4306 - accuracy: 0.8025 - auc_4: 0.8802 - auc_5: 0.9014\n",
      "1870/1870 [==============================] - 3s 1ms/step - loss: 0.6781 - accuracy: 0.6744 - auc_4: 0.7258 - auc_5: 0.7501\n",
      "\n",
      "\n",
      "Test Loss 0.6781337857246399, Test Accuracy 0.6743761301040649, Test ROC AUC 0.7258424758911133, Test PR AUC 0.7500528693199158\n",
      "Predicted good rating: 67.86%  | Actual rating label:  Bad Rating\n",
      "Predicted good rating: 22.11%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 98.32%  | Actual rating label:  Bad Rating\n",
      "Predicted good rating: 47.28%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 2.11%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 90.36%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 78.45%  | Actual rating label:  Bad Rating\n",
      "Predicted good rating: 78.14%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 50.58%  | Actual rating label:  Good Rating\n",
      "Predicted good rating: 49.05%  | Actual rating label:  Bad Rating\n",
      "Predicted good rating: 80.14%  | Actual rating label:  Bad Rating\n",
      "Predicted good rating: 96.52%  | Actual rating label:  Bad Rating\n"
     ]
    }
   ],
   "source": [
    "# load sample as tf dataset\n",
    "def get_dataset(file_path):\n",
    "    dataset = tf.data.experimental.make_csv_dataset(\n",
    "        file_path,\n",
    "        batch_size=12,\n",
    "        label_name='label',\n",
    "        na_value=\"0\",\n",
    "        num_epochs=1,\n",
    "        ignore_errors=True)\n",
    "    return dataset\n",
    "\n",
    "\n",
    "# split as test dataset and training dataset\n",
    "train_dataset = get_dataset(training_samples_file_path)\n",
    "test_dataset = get_dataset(test_samples_file_path)\n",
    "\n",
    "# movie id embedding feature\n",
    "movie_col = tf.feature_column.categorical_column_with_identity(key='movieId', num_buckets=1001)\n",
    "movie_emb_col = tf.feature_column.embedding_column(movie_col, 10)\n",
    "\n",
    "# user id embedding feature\n",
    "user_col = tf.feature_column.categorical_column_with_identity(key='userId', num_buckets=30001)\n",
    "user_emb_col = tf.feature_column.embedding_column(user_col, 10)\n",
    "\n",
    "# define input for keras model\n",
    "inputs = {\n",
    "    'movieId': tf.keras.layers.Input(name='movieId', shape=(), dtype='int32'),\n",
    "    'userId': tf.keras.layers.Input(name='userId', shape=(), dtype='int32'),\n",
    "}\n",
    "\n",
    "\n",
    "# neural cf model arch two. only embedding in each tower, then MLP as the interaction layers\n",
    "# Neural CF\n",
    "def neural_cf_model_1(feature_inputs, item_feature_columns, user_feature_columns, hidden_units):\n",
    "    item_tower = tf.keras.layers.DenseFeatures(item_feature_columns)(feature_inputs)\n",
    "    user_tower = tf.keras.layers.DenseFeatures(user_feature_columns)(feature_inputs)\n",
    "    interact_layer = tf.keras.layers.concatenate([item_tower, user_tower])\n",
    "    for num_nodes in hidden_units:\n",
    "        interact_layer = tf.keras.layers.Dense(num_nodes, activation='relu')(interact_layer)\n",
    "    output_layer = tf.keras.layers.Dense(1, activation='sigmoid')(interact_layer)\n",
    "    neural_cf_model = tf.keras.Model(feature_inputs, output_layer)\n",
    "    return neural_cf_model\n",
    "\n",
    "\n",
    "# neural cf model arch one. embedding+MLP in each tower, then dot product layer as the output\n",
    "# 双塔模型\n",
    "def neural_cf_model_2(feature_inputs, item_feature_columns, user_feature_columns, hidden_units):\n",
    "    item_tower = tf.keras.layers.DenseFeatures(item_feature_columns)(feature_inputs)\n",
    "    for num_nodes in hidden_units:\n",
    "        item_tower = tf.keras.layers.Dense(num_nodes, activation='relu')(item_tower)\n",
    "\n",
    "    user_tower = tf.keras.layers.DenseFeatures(user_feature_columns)(feature_inputs)\n",
    "    for num_nodes in hidden_units:\n",
    "        user_tower = tf.keras.layers.Dense(num_nodes, activation='relu')(user_tower)\n",
    "    # \n",
    "    output = tf.keras.layers.Dot(axes=1)([item_tower, user_tower])\n",
    "\n",
    "    neural_cf_model = tf.keras.Model(feature_inputs, output)\n",
    "    return neural_cf_model\n",
    "\n",
    "\n",
    "# neural cf model architecture\n",
    "model = neural_cf_model_1(inputs, [movie_emb_col], [user_emb_col], [10, 10])\n",
    "\n",
    "# compile the model, set loss function, optimizer and evaluation metrics\n",
    "model.compile(\n",
    "    loss='binary_crossentropy',\n",
    "    optimizer='adam',\n",
    "    metrics=['accuracy', tf.keras.metrics.AUC(curve='ROC'), tf.keras.metrics.AUC(curve='PR')])\n",
    "\n",
    "# train the model\n",
    "model.fit(train_dataset, epochs=5)\n",
    "\n",
    "# evaluate the model\n",
    "test_loss, test_accuracy, test_roc_auc, test_pr_auc = model.evaluate(test_dataset)\n",
    "print('\\n\\nTest Loss {}, Test Accuracy {}, Test ROC AUC {}, Test PR AUC {}'.format(test_loss, test_accuracy,\n",
    "                                                                                   test_roc_auc, test_pr_auc))\n",
    "\n",
    "# print some predict results\n",
    "predictions = model.predict(test_dataset)\n",
    "for prediction, goodRating in zip(predictions[:12], list(test_dataset)[0][1][:12]):\n",
    "    print(\"Predicted good rating: {:.2%}\".format(prediction[0]),\n",
    "          \" | Actual rating label: \",\n",
    "          (\"Good Rating\" if bool(goodRating) else \"Bad Rating\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf.keras.utils.plot_model(model, to_file=\"NeuralCF.png\", show_shapes=True)"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}