BoW-TfIdf-Word2Vec-Doc2Vec.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "dbb6cb41-fcc7-4dae-9881-a01bee6c1a6e",
   "metadata": {},
   "source": [
    "# **Sentiment Analysis** of the **Restaurant Reviews** from **YELP** Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cab94717-f8fa-4ec7-9cf3-b81763c74f3f",
   "metadata": {},
   "source": [
    "### - Bag-of-words and TF-IDF"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b11169b7-80ba-4d98-8164-70542c9b27ae",
   "metadata": {},
   "source": [
    "### - Word2Vec and Doc2Vec"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "38e23404-8c3d-497a-9e2e-e01e71db905c",
   "metadata": {},
   "source": [
    "#### Importing libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "f91305f6-2bce-450a-b333-ba9093ef5ded",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from gensim.models import Doc2Vec\n",
    "from gensim.models.doc2vec import TaggedDocument\n",
    "from gensim.models import Word2Vec\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import accuracy_score,f1_score,precision_score,recall_score,confusion_matrix,classification_report\n",
    "from sklearn.metrics import plot_confusion_matrix\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import svm\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "import sys\n",
    "import os\n",
    "import warnings\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f415f3ce-a0f4-4676-8728-53090c70cf68",
   "metadata": {},
   "outputs": [],
   "source": [
    "if not sys.warnoptions:\n",
    "    warnings.simplefilter(\"ignore\")\n",
    "    os.environ[\"PYTHONWARNINGS\"] = \"ignore\" "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a5c01c4b-7438-4edb-b5ec-fd09d43d088c",
   "metadata": {},
   "source": [
    "#### **Read** the **cleaned dataset stored** in a GCP **bucket after** the **preprocesssing** phase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0dd72e00-7e7b-4be1-bfe4-37707a45a2b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_final = pd.read_csv('gs://demo2222/df_final.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b83eb150-093c-4e2a-bebb-1504c65da3cd",
   "metadata": {},
   "source": [
    "#### Train/test split **keeping** the two **columns of interest** and **preserve** the **same proportions** of examples in **each class** as **observed** in the **original dataset**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "641e34c7-a65c-44f8-be38-ce0cebdd7e27",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = df_final['cleaned_reviews'] \n",
    "y = df_final['sentiment']\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3, random_state = 42, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "957a2257-602f-47ee-bfca-f6e43e52dbdc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3660    journeyman want like try hard consider restaur...\n",
       "1395    sad old chef leave start come delicious menu c...\n",
       "7620    come familiar expect parent surprised generous...\n",
       "5904    aldo server excellent attentive point come cel...\n",
       "9029    good chinese town order time count hand sesame...\n",
       "Name: cleaned_reviews, dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "fb75364d-08f0-4866-9d56-b0695e0e22ed",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6500,)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c85d30d1-7e57-4c48-a35e-57e38e109ba2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.isnull().any().any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "ff896a71-5db4-4917-9657-adbcd9d1d21f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3660    0.0\n",
       "1395    0.0\n",
       "7620    1.0\n",
       "5904    1.0\n",
       "9029    1.0\n",
       "Name: sentiment, dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6e2789fe-400d-49d7-ae6c-9e519d65c884",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6500,)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de35c823-45c6-41a9-91d2-3bcdb9e5fe4c",
   "metadata": {},
   "source": [
    "#### UDF for **evaluating** models\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ece228b9-70c5-4e41-9726-603b8a7515c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(y_test, predictions):\n",
    "    cf_matrix = confusion_matrix(y_test, predictions)\n",
    "    sns.heatmap(cf_matrix, annot = True, fmt = 'd',cmap=\"Blues\")\n",
    "    plt.title('Heatmap of confusion matrix for Test data')\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label') "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "297e9eda-f0f7-4cc3-adfc-a9d2d599004c",
   "metadata": {},
   "source": [
    "###  **Bag-of-Words**. Establishing benchmarks: Logistic Regression and SVM models with default parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "55d3d83c-1721-4396-b80e-2c8249472d00",
   "metadata": {},
   "source": [
    "#### Let's **find out** which **ngram_range performs better** with the **Logistic Regression** model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "232cd63e-8bda-4c77-9d67-bdd3223e6b9e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best ngram_range: (1, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid = {'c_vectorizer__ngram_range': [(1, 1), (2, 2), (3, 3), (1, 2), (2, 3), (1, 3)]}\n",
    "\n",
    "lr_pipeline = Pipeline([\n",
    "    ('c_vectorizer', CountVectorizer()),\n",
    "    ('lr', LogisticRegression(random_state=42))\n",
    "])\n",
    "\n",
    "\n",
    "grid_search_lr = GridSearchCV(lr_pipeline, cv=2, param_grid=param_grid, scoring='f1')\n",
    "grid_search_lr.fit(X_train, y_train)\n",
    "\n",
    "print('Best ngram_range:', grid_search_lr.best_estimator_.get_params()['c_vectorizer__ngram_range'])\n",
    "\n",
    "predictions = grid_search_lr.predict(X_test)\n",
    "evaluate(y_test, predictions )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "95138e9c-105a-4ca8-beae-33c460d27427",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9616    0.9878    0.9745      1393\n",
      "         1.0     0.9875    0.9605    0.9738      1393\n",
      "\n",
      "    accuracy                         0.9742      2786\n",
      "   macro avg     0.9745    0.9742    0.9742      2786\n",
      "weighted avg     0.9745    0.9742    0.9742      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad9036c9-19ac-43c4-b1d6-75f0a2759cf3",
   "metadata": {},
   "source": [
    "#### Let's **find out** which **ngram_range performs better** for with the **SVM** model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "b4099dee-5a3f-4567-811d-29c499a2024b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best ngram_range: (1, 2)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid = {'c_vectorizer__ngram_range': [(1, 1), (2, 2), (3, 3), (1, 2), (2, 3), (1, 3)]}\n",
    "\n",
    "svm_pipe = Pipeline([\n",
    " ('c_vectorizer', CountVectorizer()),\n",
    " ('svm', svm.SVC(max_iter=-1, random_state=123))\n",
    "])\n",
    "\n",
    "grid_search_svm = GridSearchCV(svm_pipe, cv=2, param_grid=param_grid, scoring='f1')\n",
    "grid_search_svm.fit(X_train, y_train)\n",
    "\n",
    "print('Best ngram_range:', grid_search_svm.best_estimator_.get_params()['c_vectorizer__ngram_range'])\n",
    "\n",
    "predictions = grid_search_svm.predict(X_test)\n",
    "evaluate(y_test, predictions )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "6fa45120-43a0-4fff-a5de-367688bb340f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9562    0.9727    0.9644      1393\n",
      "         1.0     0.9722    0.9555    0.9638      1393\n",
      "\n",
      "    accuracy                         0.9641      2786\n",
      "   macro avg     0.9642    0.9641    0.9641      2786\n",
      "weighted avg     0.9642    0.9641    0.9641      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e5de6930-a4c3-4cca-b442-e17415115410",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **Logistic Regression** model with **hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "df68ce29-fccc-4396-8dec-58f4fced6824",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best l1_ratio: none\n",
      "Best C: 1\n",
      "Best penalty: l2\n",
      "Best solver: lbfgs\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid = {\"lr__C\":[0.5, 1, 2], 'lr__l1_ratio':['none', 0, 0.1], 'lr__penalty':['l2', 'elasticnet'],\n",
    "       'lr__solver':['lbfgs', 'saga']}\n",
    "\n",
    "# Pipeline for CountVectorizer - with the best ngram_range - and Logistic Regression\n",
    "lr_pipe_2 = Pipeline([\n",
    " ('c_vectorizer', CountVectorizer(ngram_range = (1,3))),\n",
    " ('lr', LogisticRegression(max_iter=1500, random_state=42))\n",
    "])\n",
    "\n",
    "logreg_cv=GridSearchCV(lr_pipe_2, grid, scoring='f1', cv=2)\n",
    "logreg_cv.fit(X_train, y_train)\n",
    "\n",
    "print('Best l1_ratio:', logreg_cv.best_estimator_.get_params()['lr__l1_ratio'])\n",
    "print('Best C:', logreg_cv.best_estimator_.get_params()['lr__C'])\n",
    "print('Best penalty:', logreg_cv.best_estimator_.get_params()['lr__penalty'])\n",
    "print('Best solver:', logreg_cv.best_estimator_.get_params()['lr__solver'])\n",
    "\n",
    "predictions = logreg_cv.predict(X_test)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "d508a58f-9188-40a7-83db-8fa670c9504b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9616    0.9878    0.9745      1393\n",
      "         1.0     0.9875    0.9605    0.9738      1393\n",
      "\n",
      "    accuracy                         0.9742      2786\n",
      "   macro avg     0.9745    0.9742    0.9742      2786\n",
      "weighted avg     0.9745    0.9742    0.9742      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a80b4781-7bfa-43cd-a2bc-95d5c3de71a4",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **SVM** model with **hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "1802ccd1-5ddb-4309-bba7-9d6f4609a3a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best KERNEL: rbf\n",
      "Best C: 3\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid_svm = [\n",
    "    {'svm__kernel':('linear', 'rbf')},\n",
    "    {'svm__C':[1, 2, 3, 4]},\n",
    "]    \n",
    "\n",
    "# Pipeline for CountVectorizer - with the best ngram_range - and SVM model\n",
    "svm_pipe_2 = Pipeline([\n",
    " ('c_vectorizer', CountVectorizer(ngram_range = (1,2))),\n",
    " ('svm',svm.SVC(random_state=123))\n",
    "])\n",
    "\n",
    "svm_model_cv = GridSearchCV(svm_pipe_2, param_grid_svm)\n",
    "\n",
    "svm_model_cv.fit(X_train, y_train)\n",
    "\n",
    "print('Best KERNEL:', svm_model_cv.best_estimator_.get_params()['svm__kernel'])\n",
    "print('Best C:', svm_model_cv.best_estimator_.get_params()['svm__C'])\n",
    "\n",
    "\n",
    "predictions = svm_model_cv.predict(X_test)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "296450fb-9f9e-4763-bd1b-2275283e858f",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9661    0.9821    0.9740      1393\n",
      "         1.0     0.9818    0.9655    0.9736      1393\n",
      "\n",
      "    accuracy                         0.9738      2786\n",
      "   macro avg     0.9739    0.9738    0.9738      2786\n",
      "weighted avg     0.9739    0.9738    0.9738      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf457371-547f-4bb1-a24c-78105d4cb9b9",
   "metadata": {},
   "source": [
    "###  TF-IDF. Establishing benchmarks: Logistic Regression and SVM models with **default parameters**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b1b0266-52d1-4bf4-9fe6-4a4dff089922",
   "metadata": {},
   "source": [
    "#### Let's **find out** which **ngram_range performs better** for the **Logistic Regression**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "efaddeb8-4a55-4257-81c6-3da3d1b9d59a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best ngram_range: (1, 2)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid = {'tfidf__ngram_range': [(1, 1), (2, 2), (3, 3), (1, 2), (2, 3), (1, 3)]}\n",
    "\n",
    "lr_pipe_tfidf = Pipeline([\n",
    "    ('tfidf', TfidfVectorizer()),\n",
    "    ('lr', LogisticRegression(max_iter=1500, random_state=42))\n",
    "])\n",
    "\n",
    "\n",
    "logreg_cv_2 = GridSearchCV(lr_pipe_tfidf, cv=2, param_grid=param_grid, scoring='f1')\n",
    "logreg_cv_2.fit(X_train, y_train)\n",
    "\n",
    "print('Best ngram_range:', logreg_cv_2.best_estimator_.get_params()['tfidf__ngram_range'])\n",
    "\n",
    "predictions = logreg_cv_2.predict(X_test)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ef2a863d-b1a5-46ae-86b4-2a428aaed90d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9448    0.9706    0.9575      1393\n",
      "         1.0     0.9697    0.9433    0.9563      1393\n",
      "\n",
      "    accuracy                         0.9569      2786\n",
      "   macro avg     0.9573    0.9569    0.9569      2786\n",
      "weighted avg     0.9573    0.9569    0.9569      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0877af7-a252-4f4c-ba79-5aca80813c83",
   "metadata": {},
   "source": [
    "#### Let's **find out** which **ngram_range performs better** for with the **SVM** model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "ea6b51c8-d550-4cad-b400-8fa494aa20d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best ngram_range: (1, 1)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid = {'tfidf__ngram_range': [(1, 1), (2, 2), (3, 3), (1, 2), (2, 3), (1, 3)]}\n",
    "\n",
    "svm_pipe_tfidf = Pipeline([\n",
    "    ('tfidf', TfidfVectorizer()),\n",
    "    ('svm', svm.SVC(max_iter=-1, random_state=123)),\n",
    "])\n",
    "\n",
    "svm_model_cv_2 = GridSearchCV(svm_pipe_tfidf, cv=2, param_grid=param_grid, scoring='f1')\n",
    "svm_model_cv_2.fit(X_train, y_train)\n",
    "\n",
    "print('Best ngram_range:', svm_model_cv_2.best_estimator_.get_params()['tfidf__ngram_range'])\n",
    "\n",
    "predictions = svm_model_cv_2.predict(X_test)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "31084022-1a4c-448a-90f8-cec6e3fea3ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9730    0.9842    0.9786      1393\n",
      "         1.0     0.9840    0.9727    0.9783      1393\n",
      "\n",
      "    accuracy                         0.9785      2786\n",
      "   macro avg     0.9785    0.9785    0.9785      2786\n",
      "weighted avg     0.9785    0.9785    0.9785      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d42a75cf-56a0-4200-8141-ffe458aa8664",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **Logistic Regression** model with **hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "09c3b79b-cdbe-4cff-8ee4-6a31e7a5e181",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best l1_ratio: none\n",
      "Best C: 30\n",
      "Best penalty: l2\n",
      "Best solver: saga\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid = {\"lr__C\":[1, 10, 20, 30, 40], 'lr__l1_ratio':['none', 0, 0.1], 'lr__penalty':['l2', 'elasticnet'],\n",
    "       'lr__solver':['lbfgs', 'saga']}\n",
    "\n",
    "# Pipeline for TfidfVectorizer - with the best ngram_range - and Logistic Regression Classifier \n",
    "lr_pipe_tfidf_2 = Pipeline([\n",
    " ('tfidf', TfidfVectorizer(ngram_range = (1,2))),\n",
    " ('lr', LogisticRegression(max_iter=1500, n_jobs = -1, random_state=42))\n",
    "])\n",
    "\n",
    "logreg_cv_3=GridSearchCV(lr_pipe_tfidf_2, grid, scoring='f1', cv=2)\n",
    "logreg_cv_3.fit(X_train, y_train)\n",
    "\n",
    "print('Best l1_ratio:', logreg_cv_3.best_estimator_.get_params()['lr__l1_ratio'])\n",
    "print('Best C:', logreg_cv_3.best_estimator_.get_params()['lr__C'])\n",
    "print('Best penalty:', logreg_cv_3.best_estimator_.get_params()['lr__penalty'])\n",
    "print('Best solver:', logreg_cv_3.best_estimator_.get_params()['lr__solver'])\n",
    "\n",
    "predictions = logreg_cv_3.predict(X_test)\n",
    "evaluate(y_test, predictions )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "6790881f-638a-448e-80d2-5251e02fe3cf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9617    0.9907    0.9760      1393\n",
      "         1.0     0.9904    0.9605    0.9752      1393\n",
      "\n",
      "    accuracy                         0.9756      2786\n",
      "   macro avg     0.9760    0.9756    0.9756      2786\n",
      "weighted avg     0.9760    0.9756    0.9756      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86f543a5-30eb-4a69-a7e4-b65d7a9feb6e",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **SVM** model with **hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b7f4372a-7e88-48d5-991a-cdbf889cc4e1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best KERNEL: rbf\n",
      "Best C: 1.5\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid_svm = [\n",
    "    {'svm__kernel':('linear', 'rbf')},\n",
    "    {'svm__C':[0.5, 1, 1.5, 1.9]},\n",
    "]    \n",
    "\n",
    "# Pipeline for TfidfVectorizer - with the best ngram_range - and SVM Classifier \n",
    "\n",
    "svm_pipe_tfidf_2 = Pipeline([\n",
    " ('tfidf', TfidfVectorizer(ngram_range = (1,1))),\n",
    " ('svm',svm.SVC(random_state=123))\n",
    "])\n",
    "\n",
    "svm_model_cv_3 = GridSearchCV(svm_pipe_tfidf_2, param_grid_svm, cv=2)\n",
    "\n",
    "svm_model_cv_3.fit(X_train, y_train)\n",
    "\n",
    "print('Best KERNEL:', svm_model_cv_3.best_estimator_.get_params()['svm__kernel'])\n",
    "print('Best C:', svm_model_cv_3.best_estimator_.get_params()['svm__C'])\n",
    "\n",
    "predictions = svm_model_cv_3.predict(X_test)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "7961d07c-e1d7-430a-a9bb-556af6ac5e19",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9745    0.9864    0.9804      1393\n",
      "         1.0     0.9862    0.9742    0.9801      1393\n",
      "\n",
      "    accuracy                         0.9803      2786\n",
      "   macro avg     0.9803    0.9803    0.9803      2786\n",
      "weighted avg     0.9803    0.9803    0.9803      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee1b5227-0743-4b1c-a344-19f61fca8d0c",
   "metadata": {},
   "source": [
    "### Word2Vec Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f446344e-f744-42b5-b0e6-a685d3759b7a",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_cleaned_reviews = X_train.str.split() # tokenization\n",
    "test_cleaned_reviews = X_test.str.split() # tokenization"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8788a7e6-96f9-4be4-9dd6-5b165de42372",
   "metadata": {},
   "source": [
    "#### **Although our dataset is not very large**, we will use **Word2Vec without using pretrained word embeddings** because we think we **have enough data to go for a custom vectorization** as it **will be very specific to the context the corpus has**." 
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34623588-1f57-4148-9dd7-8925133f9e0c",
   "metadata": {},
   "source": [
    "#### Word2Vec model with **default parameters** except for the number of features (500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "8b2ffe3d-cdff-4998-9f15-f2da77b82a67",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5047666, 5926340)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "number_of_features = 500\n",
    "model_w2v = Word2Vec(\n",
    "            train_cleaned_reviews,\n",
    "            size=number_of_features, # nmumber of features\n",
    "            window=5, # window size\n",
    "            min_count=5, # Ignores all words with total frequency lower than 5.                                  \n",
    "            sg = 1, # 1 for skip-gram model\n",
    "            hs = 0, # negative sampling will be used\n",
    "            negative = 10, # > 0, for negative sampling\n",
    "            workers= 4, # number of cores\n",
    "            seed = 42 \n",
    ") \n",
    "\n",
    "model_w2v.train(train_cleaned_reviews, total_examples= len(X_train), epochs=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0befd030-5fc2-4ecd-9670-239a7526a0d5",
   "metadata": {},
   "source": [
    "#### UDF to **get the vectors** that will **feed** the **classifiers**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "42860793-c8db-43f4-82f7-88563a278d60",
   "metadata": {},
   "outputs": [],
   "source": [
    "def getVectors(dataset):\n",
    "  word_Embedding=np.zeros(number_of_features) \n",
    "  vectors=[]\n",
    "  for review in dataset:\n",
    "    wordCount=0\n",
    "    # Loop over each word in the review and, if it is in the model's vocaublary, add its feature vector to the total\n",
    "    for word in review: \n",
    "      if word in model_w2v.wv.vocab: # model_w2v.wv.vocab is a list that contains the names of the words in the model's vocabulary\n",
    "        word_Embedding=word_Embedding+model_w2v.wv[word] \n",
    "        wordCount=wordCount+1\n",
    "  \n",
    "    word_Embedding=word_Embedding/wordCount  # Get the average\n",
    "    vectors.append(word_Embedding)\n",
    "  return vectors\n",
    "\n",
    "trainReviewVectors=getVectors(train_cleaned_reviews)\n",
    "testReviewVectors=getVectors(test_cleaned_reviews)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35b7eed6-e547-4bbe-9d6f-ab1fc0933f2d",
   "metadata": {},
   "source": [
    "#### Establishing **benchmarks** with two **classifiers** with the **default hyperparameters**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a1ba7538-84ac-4502-a576-462a67c6bb14",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **Logistic Regression** with the **default hyperparameters**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "dda73d43-28fe-4366-838d-7cb87c631a6e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_lr = LogisticRegression(random_state=42)\n",
    "model_lr.fit(trainReviewVectors, y_train)\n",
    "predictions = model_lr.predict(testReviewVectors)\n",
    "evaluate(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "d3a10b6f-2803-47f2-a6b5-d22262b259fb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9224    0.9124    0.9174      1393\n",
      "         1.0     0.9134    0.9232    0.9182      1393\n",
      "\n",
      "    accuracy                         0.9178      2786\n",
      "   macro avg     0.9179    0.9178    0.9178      2786\n",
      "weighted avg     0.9179    0.9178    0.9178      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e1e72fe-a58e-4e95-ba43-ca1d668e18e3",
   "metadata": {},
   "source": [
    "#### **Evaluation** using **SVM** with the **default hyperparameters**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "dedd4412-b114-42fb-8bde-845dbe39c992",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_svm = svm.SVC(random_state=42)\n",
    "model_svm.fit(trainReviewVectors, y_train)\n",
    "predictions = model_svm.predict(testReviewVectors)\n",
    "evaluate(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "b9f1a9fb-8d4e-4807-9539-4a6c6f5dd2f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9285    0.9232    0.9258      1393\n",
      "         1.0     0.9236    0.9289    0.9263      1393\n",
      "\n",
      "    accuracy                         0.9261      2786\n",
      "   macro avg     0.9261    0.9261    0.9261      2786\n",
      "weighted avg     0.9261    0.9261    0.9261      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9d9da7d-80fe-40b5-b28e-140e0f4c7e0c",
   "metadata": {},
   "source": [
    "#### **Logistic Regression** with **Hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "818e241f-b6b8-408f-9eb5-9eb5087d2d25",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best C: 30\n",
      "Best l1_ratio: none\n",
      "Best penalty: l2\n",
      "Best solver: lbfgs\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid_lr_word2vec = {\"C\":[1, 10, 20, 30, 50], 'l1_ratio':['none', 0, 0.1], 'penalty':['l2', 'elasticnet'],\n",
    "                'solver':['lbfgs', 'saga']}\n",
    "\n",
    "lr_word2vec = LogisticRegression(max_iter=1500, random_state=42)\n",
    "\n",
    "logreg_cv_word2vec=GridSearchCV(lr_word2vec, grid_lr_word2vec, n_jobs=-1)\n",
    "logreg_cv_word2vec.fit(trainReviewVectors, y_train)\n",
    "\n",
    "print('Best C:', logreg_cv_word2vec.best_estimator_.get_params()['C'])\n",
    "print('Best l1_ratio:', logreg_cv_word2vec.best_estimator_.get_params()['l1_ratio'])\n",
    "print('Best penalty:', logreg_cv_word2vec.best_estimator_.get_params()['penalty'])\n",
    "print('Best solver:', logreg_cv_word2vec.best_estimator_.get_params()['solver'])\n",
    "\n",
    "\n",
    "predictions = logreg_cv_word2vec.predict(testReviewVectors)\n",
    "evaluate(y_test, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "389d16a2-814f-472e-99dd-ad415e5f18aa",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9224    0.9296    0.9260      1393\n",
      "         1.0     0.9291    0.9218    0.9254      1393\n",
      "\n",
      "    accuracy                         0.9257      2786\n",
      "   macro avg     0.9257    0.9257    0.9257      2786\n",
      "weighted avg     0.9257    0.9257    0.9257      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9fdebd55-488c-45af-a59c-bda8043d424b",
   "metadata": {},
   "source": [
    "#### **SVM** with **Hyperparameter tuning** (using **GridSearchCV**)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "bd10cf3a-a34f-4846-ae85-d4edc068e7f0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best KERNEL: rbf\n",
      "Best C: 30\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Hyperparameter tuning with SVM \n",
    "\n",
    "grid_svm_word2vec = [\n",
    "    {'kernel':('linear', 'rbf')},\n",
    "    {'C':[1, 10, 20, 25, 30, 35]},\n",
    "]    \n",
    "\n",
    "svm_word2vec = svm.SVC(random_state=42)\n",
    "\n",
    "\n",
    "svm_word2vec_cv = GridSearchCV(svm_word2vec, grid_svm_word2vec, n_jobs=-1)\n",
    "\n",
    "svm_word2vec_cv.fit(trainReviewVectors, y_train)\n",
    "\n",
    "print('Best KERNEL:', svm_word2vec_cv.best_estimator_.get_params()['kernel'])\n",
    "print('Best C:', svm_word2vec_cv.best_estimator_.get_params()['C'])\n",
    "\n",
    "predictions = svm_word2vec_cv.predict(testReviewVectors)\n",
    "evaluate(y_test, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "602075b0-9f53-44ee-8dc7-a7b994462471",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9527    0.9842    0.9682      1393\n",
      "         1.0     0.9837    0.9512    0.9672      1393\n",
      "\n",
      "    accuracy                         0.9677      2786\n",
      "   macro avg     0.9682    0.9677    0.9677      2786\n",
      "weighted avg     0.9682    0.9677    0.9677      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cef7b232-1aec-401e-9f53-41fd5565b30d",
   "metadata": {},
   "source": [
    "### **Doc2Vec** model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9de51985-c639-4130-834e-9387b4ed9355",
   "metadata": {},
   "source": [
    "#### To keep **features** and **labels** in only **two** datasets, let's concatenate X_train and y_train into a **train set**, and X_test and y_test into a **test set**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "5f4a42c9-7a16-484e-bc87-602fe30bcb35",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6500, 2)"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train = pd.concat([X_train, y_train], axis=1)\n",
    "train = train.reset_index(drop=True)\n",
    "test = pd.concat([X_test, y_test], axis=1)\n",
    "test = test.reset_index(drop=True)\n",
    "train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "2852ef79-c428-46e2-906f-8ce29f5067fc",
   "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>cleaned_reviews</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>journeyman want like try hard consider restaur...</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>sad old chef leave start come delicious menu c...</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>come familiar expect parent surprised generous...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>aldo server excellent attentive point come cel...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>good chinese town order time count hand sesame...</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     cleaned_reviews  sentiment\n",
       "0  journeyman want like try hard consider restaur...        0.0\n",
       "1  sad old chef leave start come delicious menu c...        0.0\n",
       "2  come familiar expect parent surprised generous...        1.0\n",
       "3  aldo server excellent attentive point come cel...        1.0\n",
       "4  good chinese town order time count hand sesame...        1.0"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "74d24f4f-6ba5-490a-b815-b6e33765f942",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['cleaned_reviews', 'sentiment'], dtype='object')"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "22824f23-a2ba-4bbe-a4c4-e92e148f92cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_tagged =  [TaggedDocument((train['cleaned_reviews'].str.split(' ')[i]), [train['sentiment'].loc[i]]) for i in range(len(train))]\n",
    "test_tagged =  [TaggedDocument((test['cleaned_reviews'].str.split(' ')[i]), [test['sentiment'].loc[i]]) for i in range(len(test))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "b31c4af5-96e2-437c-ba67-dfc604b5cc10",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TaggedDocument(words=['journeyman', 'want', 'like', 'try', 'hard', 'consider', 'restaurant', 'literally', 'minute', 'house', 'neighbor', 'love', 'union', 'square', 'refinement', 'journeyman', 'purportedly', 'bring', 'definitely', 'overdue', 'work', 'paper', 'restaurant', 'love', 'local', 'modern', 'inventive', 'casual', 'reality', 'restaurant', 'seriously', 'despite', 'disappointing', 'dining', 'experience', 'glad', 'return', 'hub', 'celebrate', 'wedding', 'anniversary', 'opt', 'course', 'omnivore', 'menu', 'add', 'serving', 'foie', 'gra', 'confit', 'veal', 'tongue', 'smoke', 'beef', 'heart', 'course', 'menu', 'consist', 'spring', 'salad', 'favas', 'pea', 'spinach', 'carrot', 'spring', 'foam', 'serve', 'purpose', 'asparagus', 'soy', 'bonito', 'sous', 'vide', 'black', 'cod', 'leek', 'puree', 'potato', 'veal', 'breast', 'sweetbread', 'artichoke', 'dessert', 'cane', 'sugar', 'include', 'macaron', 'shoofly', 'pie', 'different', 'thing', 'amuse', 'offer', 'warm', 'lentil', 'salad', 'gruyere', 'gougere', 'house', 'cure', 'ham', 'pre', 'dessert', 'flan', 'freshly', 'grate', 'nutmeg', 'post', 'dessert', 'poppyseed', 'cracker', 'rhubarb', 'jam', 'florentines', 'house', 'marshmallow', 'end', 'different', 'thing', 'taste', 'order', 'ask', 'cod', 'sweetbread', 'follow', 'flan', 'unfortunately', 'rest', 'discordant', 'flavor', 'harmony', 'especially', 'asparagus', 'dish', 'bland', 'surprising', 'consider', 'component', 'plate', 'plate', 'pretty', 'gel', 'foam', 'pearl', 'far', 'flavor', 'par', 'drink', 'wonderful', 'albeit', 'tiny', 'cocktail', 'like', 'oz', 'atmosphere', 'awkward', 'arrive', 'early', 'o', 'clock', 'reservation', 'immediately', 'feel', 'like', 'intruder', 'no', 'music', 'automatically', 'feel', 'like', 'whisper', 'space', 'spare', 'music', 'definitely', 'compliment', 'industrial', 'slightly', 'clinical', 'vibe', 'want', 'restaurant', 'feel', 'fun', 'fun', 'hospital', 'recovery', 'room', 'service', 'adorable', 'way', 'good', 'willed', 'effort', 'little', 'actual', 'success', 'seriously', 'think', 'server', 'explode', 'awkwardness', 'stilted', 'service', 'giggly', 'turn', 'worsen', 'awkwardness', 'ah', 'tab', 'come', 'kind', 'absurd', 'despite', 'fact', 'drink', 'half', 'substantial', 'meal', 'e', 'g', 'island', 'creek', 'oyster', 'bar', 'coppa', 'mention', 'laugh'], tags=[0.0])"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_tagged[0] # example with the first review of the train set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c19cf8d-f28c-452e-ac34-85b1f8869bd2",
   "metadata": {},
   "source": [
    "#### Doc2Vec model using **negative smapling** with **default hyperparameters** except for the number of features (500) given the **vast corpus** "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "ac3e54ea-6fa8-4deb-a772-c1d66263e076",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_dbow = Doc2Vec(dm=1, vector_size=500, workers=4, hs=0, negative=5, seed=123)\n",
    "model_dbow.build_vocab(train_tagged)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "11c27076-2dd8-480f-b790-55182dac4b85",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_dbow.train(train_tagged, total_examples=len(train_tagged), epochs=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ef9ba1c-a94f-408d-8ac4-ae53f24a9011",
   "metadata": {},
   "source": [
    "#### Building the **Final Vector Feature** for the classifiers\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "3a352388-fc3d-4281-8458-01eb20183697",
   "metadata": {},
   "outputs": [],
   "source": [
    "def vec_for_learning(model, tagged_docs):\n",
    "    sents = tagged_docs\n",
    "    targets, regressors = zip(*[(doc.tags[0], model.infer_vector(list(doc.words))) for doc in sents])\n",
    "    return targets, regressors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "a6df79d1-b050-4e52-bcc7-4024d043580d",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train_tagged, X_train_tagged = vec_for_learning(model_dbow, train_tagged)\n",
    "y_test_tagged, X_test_tagged = vec_for_learning(model_dbow, test_tagged)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "728117dd-b3a3-4967-8244-21af8c6cc879",
   "metadata": {},
   "source": [
    "#### Using the **Logistic Regression** model with the the **default hyperparameters**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "214056a5-b58e-4f0c-b122-22adde6cd704",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_lr = LogisticRegression(max_iter=100, random_state=123)\n",
    "model_lr.fit(X_train_tagged, y_train_tagged)\n",
    "predictions = model_lr.predict(X_test_tagged)\n",
    "evaluate(y_test_tagged, predictions )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "9a2dafbb-3c55-4617-bf12-b60b394f8984",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9051    0.8973    0.9012      1393\n",
      "         1.0     0.8982    0.9060    0.9021      1393\n",
      "\n",
      "    accuracy                         0.9017      2786\n",
      "   macro avg     0.9017    0.9017    0.9016      2786\n",
      "weighted avg     0.9017    0.9017    0.9016      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test_tagged, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "243bcfb3-78fa-4506-bca8-4ba7447f562c",
   "metadata": {},
   "source": [
    "#### Using **SVM** model with the the **default hyperparameters**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "c058241c-b40c-4151-9c03-5ef5dcaa42e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_svm = svm.SVC(max_iter=-1, random_state=42) \n",
    "model_svm.fit(X_train_tagged, y_train_tagged)\n",
    "predictions = model_svm.predict(X_test_tagged)\n",
    "evaluate(y_test_tagged, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "a3f6ae2e-ae6c-4900-a041-ce5dcfc8deb9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9019    0.9045    0.9032      1393\n",
      "         1.0     0.9042    0.9017    0.9029      1393\n",
      "\n",
      "    accuracy                         0.9031      2786\n",
      "   macro avg     0.9031    0.9031    0.9031      2786\n",
      "weighted avg     0.9031    0.9031    0.9031      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test_tagged, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c3f507a7-2aff-47a8-897f-a97142719de1",
   "metadata": {},
   "source": [
    "#### **Hyperparemeter tuning** (using **GridSearchCV**) for **Logistic Regression**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "9da48213-08d6-4d89-8d03-707b0cf7e3b9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best C: 1.7\n",
      "Best l1_ratio: none\n",
      "Best penalty l2\n",
      "Best solver: lbfgs\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid_doc2vec_lr = {\"C\":[1, 1.5, 1.7, 1.8], 'l1_ratio':['none', 0, 0.1], 'penalty':['l2', 'elasticnet'],\n",
    "       'solver':['lbfgs', 'saga']}\n",
    "\n",
    "model_lr = LogisticRegression(max_iter=100, random_state=123) \n",
    "logreg_cv_doc2vec=GridSearchCV(model_lr, grid_doc2vec_lr, cv=2, n_jobs=-1)\n",
    "logreg_cv_doc2vec.fit(X_train_tagged, y_train_tagged)\n",
    "\n",
    "print('Best C:', logreg_cv_doc2vec.best_estimator_.get_params()['C'])\n",
    "print('Best l1_ratio:', logreg_cv_doc2vec.best_estimator_.get_params()['l1_ratio'])\n",
    "print('Best penalty', logreg_cv_doc2vec.best_estimator_.get_params()['penalty'])\n",
    "print('Best solver:', logreg_cv_doc2vec.best_estimator_.get_params()['solver'])\n",
    "\n",
    "\n",
    "predictions = logreg_cv_doc2vec.predict(X_test_tagged)\n",
    "evaluate(y_test_tagged, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "224e8fb0-7159-4fed-8b66-8fda757e7d5e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9053    0.8988    0.9020      1393\n",
      "         1.0     0.8995    0.9060    0.9027      1393\n",
      "\n",
      "    accuracy                         0.9024      2786\n",
      "   macro avg     0.9024    0.9024    0.9024      2786\n",
      "weighted avg     0.9024    0.9024    0.9024      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test_tagged, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1970cb7e-ec5a-4adc-bf01-03a0b3057fc7",
   "metadata": {},
   "source": [
    "#### **Hyperparameter tuning** (using **GridSearchCV**) for **SVM**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "193ee1fc-e324-44a8-a143-6f1c5a5a2bc5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best KERNEL: rbf\n",
      "Best C: 8\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid_doc2vec_svm = [\n",
    "    {'kernel':('linear', 'rbf')},\n",
    "    {'C':[ 1, 4, 8, 12, 16, 20]},\n",
    "]    \n",
    "\n",
    "svm_model = svm.SVC(random_state=42)\n",
    "\n",
    "\n",
    "svm_model_cv_doc2vec = GridSearchCV(svm_model, grid_doc2vec_svm, n_jobs=-1)\n",
    "\n",
    "svm_model_cv_doc2vec.fit(X_train_tagged, y_train_tagged)\n",
    "\n",
    "print('Best KERNEL:', svm_model_cv_doc2vec.best_estimator_.get_params()['kernel'])\n",
    "print('Best C:', svm_model_cv_doc2vec.best_estimator_.get_params()['C'])\n",
    "\n",
    "predictions = svm_model_cv_doc2vec.predict(X_test_tagged)\n",
    "evaluate(y_test_tagged, predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "35a21ebc-8ac4-437f-81d8-ec8b1ddf2902",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         0.0     0.9005    0.9160    0.9082      1393\n",
      "         1.0     0.9145    0.8988    0.9066      1393\n",
      "\n",
      "    accuracy                         0.9074      2786\n",
      "   macro avg     0.9075    0.9074    0.9074      2786\n",
      "weighted avg     0.9075    0.9074    0.9074      2786\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test_tagged, predictions, digits=4))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7170dbbe-ace8-417b-af69-02f08df98c71",
   "metadata": {},
   "source": [
    "#### **Conclusions**: \n",
    "#### - For **all** models, **F1 scores** have **high values** (> 90%).\n",
    "#### - **All tuned** models have **higher values** for **F1 score** than the **corresponding non-tuned** models, as **expected** (with the exception of Logistic Regression with CountVectorizer, we got the same score for the non-tuned model and for the tuned model) .\n",
    "#### - The **best F1 scores** (for both labels) were achieved using **TF-IDF** with the **SVM** model with **hyperparameter tuning** (just above 98%). "
   ]
  }
 ],
 "metadata": {
  "environment": {
   "kernel": "python3",
   "name": "tf2-gpu.2-6.m91",
   "type": "gcloud",
   "uri": "gcr.io/deeplearning-platform-release/tf2-gpu.2-6:m91"
  },
  "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": "3.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}