Project_Sentiment_Analysis_Thanapoom.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {
    "id": "WGv2X4rTbo-U"
   },
   "source": [
    "# Project: Sentiment Analysis of Bank Customer Reviews using Bidirectional Encoder Representations from Transformers\n",
    "\n",
    "## Basic information\n",
    "- Course: CS583 - Deep Learning\n",
    "- Student name: Thanapoom Phatthanaphan\n",
    "- Student CWID: 20011296\n",
    "\n",
    "## Introduction\n",
    "In the ever-evolving landscape of the banking industry, the invaluable role of customer feedback cannot be overstated. In an era dominated by digital communication, customer reviews and textual comments wield a considerable influ-ence over consumer choices, shaping the overall reputation and performance of financial institutions. The multifaceted nature of language and the sheer volume of textual data, however, present a formidable challenge when attempting to extract meaningful insights from these reviews.\n",
    "This research embarks on a pioneering journey to navigate this challenge by adopting a sophisticated approach to sen-timent analysis. The traditional methods of sentiment analysis fall short when faced with the complexity and nu-ance inherent in customer reviews. To overcome these limi-tations, we embrace cutting-edge technologies, specifically Bidirectional Encoder Representations from Transformers (BERT) with PyTorch.\n",
    "The central objective of this research is to construct an ad-vanced sentiment analysis system that goes beyond mere classification of customer reviews as positive or negative. We delve into the nuances of customer sentiments and preferences, aiming to unravel the intricate layers of feed-back intricately woven into the vast tapestry of textual da-ta. This ambitious project focuses on a substantial dataset comprising 10,000 recent customer reviews sourced from 48 diverse banks across the United States.\n",
    "The integration of BERT embeddings with PyTorch signi-fies a departure from conventional sentiment analysis methods. BERT, known for its contextual understanding of language, enhances the model's ability to capture subtle-ties within the reviews, surpassing the limitations of tradi-tional methods. The primary goal is not merely to catego-rize reviews but to provide the banking industry with nu-anced and precise insights into customer sentiments. By leveraging the power of BERT embeddings with PyTorch, we aim to overcome the challenges inherent in textual feedback analysis and elevate the analysis to a level where it becomes a strategic tool for enhancing service quality. This research aspires to empower the banking industry, of-fering actionable insights that transcend the binary classifi-cation of positive or negative sentiments. The vision ex-tends beyond classification; it encompasses a comprehen-sive understanding of the underlying sentiments and prefer-ences expressed by customers. Ultimately, this endeavor seeks to catalyze a paradigm shift in how the banking in-dustry interprets and responds to customer feedback, foster-ing an environment of continuous improvement and heightened customer satisfaction.\n",
    "\n",
    "## Dataset\n",
    "The dataset utilized in this research comprises over 10,000 customer reviews sourced from 48 distinct banking institu- tions within the United States. This publicly available dataset can be accessed through the Kaggle community platform, offering a comprehensive collection of customer sentiments towards various banks. The dataset's primary objective is to serve as a resource for training and evaluating the sentiment analysis system proposed in this research. For easy accessibility, the dataset can be downloaded using the following link: https://www.kaggle.com/datasets/trainingdatapro/20000-customers-reviews-on-banks/data. The richness of this dataset not only allows for a detailed exploration of customer feedback but also ensures a diverse and representative sample, crucial for the effectiveness and generalizability of the sentiment analysis model.\n",
    "\n",
    "## Goal\n",
    "Building the sentiment analysis using BERT to analyze each customer review to judge that it is positive or negative review, considering 4-5 stars as positive reviews and 1-3 stars as negative reviews."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 12376,
     "status": "ok",
     "timestamp": 1701409888919,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "CXQ6VWK920jB",
    "outputId": "fd7216d8-9e6c-4a9e-f495-7cf06bf2b6bf"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n",
      "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: transformers in /usr/local/lib/python3.10/dist-packages (4.35.2)\n",
      "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from transformers) (3.13.1)\n",
      "Requirement already satisfied: huggingface-hub<1.0,>=0.16.4 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.19.4)\n",
      "Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (1.23.5)\n",
      "Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from transformers) (23.2)\n",
      "Requirement already satisfied: pyyaml>=5.1 in /usr/local/lib/python3.10/dist-packages (from transformers) (6.0.1)\n",
      "Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers) (2023.6.3)\n",
      "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers) (2.31.0)\n",
      "Requirement already satisfied: tokenizers<0.19,>=0.14 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.15.0)\n",
      "Requirement already satisfied: safetensors>=0.3.1 in /usr/local/lib/python3.10/dist-packages (from transformers) (0.4.0)\n",
      "Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers) (4.66.1)\n",
      "Requirement already satisfied: fsspec>=2023.5.0 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers) (2023.6.0)\n",
      "Requirement already satisfied: typing-extensions>=3.7.4.3 in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.16.4->transformers) (4.5.0)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (3.3.2)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (3.4)\n",
      "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2.0.7)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers) (2023.7.22)\n",
      "Requirement already satisfied: torch in /usr/local/lib/python3.10/dist-packages (2.1.0+cu118)\n",
      "Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch) (3.13.1)\n",
      "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch) (4.5.0)\n",
      "Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch) (1.12)\n",
      "Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch) (3.2.1)\n",
      "Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch) (3.1.2)\n",
      "Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from torch) (2023.6.0)\n",
      "Requirement already satisfied: triton==2.1.0 in /usr/local/lib/python3.10/dist-packages (from torch) (2.1.0)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch) (2.1.3)\n",
      "Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch) (1.3.0)\n"
     ]
    }
   ],
   "source": [
    "# Import the necessary libraries for performing data analysis\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# import necessary libraries for cleaning text\n",
    "import nltk\n",
    "nltk.download('stopwords')\n",
    "nltk.download('punkt')\n",
    "from nltk.corpus import stopwords\n",
    "import re\n",
    "\n",
    "# Install the required libraries for BERT\n",
    "!pip install transformers\n",
    "!pip install torch\n",
    "\n",
    "# Import the necessary libraries for BERT\n",
    "import torch\n",
    "import torch.nn.functional as F\n",
    "from transformers import BertTokenizer, BertForSequenceClassification, AdamW\n",
    "from torch.utils.data import DataLoader, TensorDataset\n",
    "from sklearn.model_selection import train_test_split\n",
    "import time\n",
    "from prettytable import PrettyTable\n",
    "\n",
    "# Import the necessary libraries for evaluation\n",
    "from sklearn.metrics import roc_curve, roc_auc_score, accuracy_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d20VJjpS22lA"
   },
   "source": [
    "## 1. Importing the dataset\n",
    "Import the dataset of customer reviews from 48 distinct banks across the United States."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1148,
     "status": "ok",
     "timestamp": 1701409890062,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "BcW0fDBt3Zzr",
    "outputId": "9f9f78d2-698e-4972-d953-4fe1db77ca21"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
     ]
    }
   ],
   "source": [
    "# Allow access to google drive where\n",
    "from google.colab import drive\n",
    "drive.mount('/content/drive')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 424
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1701409890062,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "IG5_lwhd3J6A",
    "outputId": "976d113c-c122-4e42-ba3b-d2f10dc021c2"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-3a4524b1-8390-4758-af4b-6f84c694af69\" class=\"colab-df-container\">\n",
       "    <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>author</th>\n",
       "      <th>date</th>\n",
       "      <th>location</th>\n",
       "      <th>bank</th>\n",
       "      <th>star</th>\n",
       "      <th>text</th>\n",
       "      <th>like</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Kyle</td>\n",
       "      <td>31.08.2023</td>\n",
       "      <td>Magnolia, TX</td>\n",
       "      <td>merrick_bank</td>\n",
       "      <td>5</td>\n",
       "      <td>Very easy to use to view statements and make o...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Julicia</td>\n",
       "      <td>23.08.2023</td>\n",
       "      <td>Columbus, GA</td>\n",
       "      <td>merrick_bank</td>\n",
       "      <td>5</td>\n",
       "      <td>Merrick Bank has always been good to me for bu...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Karen</td>\n",
       "      <td>2.06.2023</td>\n",
       "      <td>Marrero, LA</td>\n",
       "      <td>merrick_bank</td>\n",
       "      <td>4</td>\n",
       "      <td>Times are tough for everyone and I have worked...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Brent</td>\n",
       "      <td>29.03.2023</td>\n",
       "      <td>Moultrie, GA</td>\n",
       "      <td>merrick_bank</td>\n",
       "      <td>5</td>\n",
       "      <td>I can not asked for a better Credit Card Compa...</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Sharon</td>\n",
       "      <td>23.11.2022</td>\n",
       "      <td>Burnham, IL</td>\n",
       "      <td>merrick_bank</td>\n",
       "      <td>5</td>\n",
       "      <td>Updated on 02/10/2023: I was happy to sign for...</td>\n",
       "      <td>3.0</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>19266</th>\n",
       "      <td>J.</td>\n",
       "      <td>30.01.2017</td>\n",
       "      <td>Salem, OR</td>\n",
       "      <td>tcf_bank</td>\n",
       "      <td>1</td>\n",
       "      <td>Paid my 1st payment on time. They sent me a la...</td>\n",
       "      <td>11.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19267</th>\n",
       "      <td>Destiny</td>\n",
       "      <td>28.01.2017</td>\n",
       "      <td>Andover, MN</td>\n",
       "      <td>tcf_bank</td>\n",
       "      <td>1</td>\n",
       "      <td>I have banked with TCF for about 4 years now a...</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19268</th>\n",
       "      <td>Sean</td>\n",
       "      <td>25.01.2017</td>\n",
       "      <td>Bothell, WA</td>\n",
       "      <td>tcf_bank</td>\n",
       "      <td>1</td>\n",
       "      <td>Most inconvenient bank ever. As a business own...</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19269</th>\n",
       "      <td>Edgar</td>\n",
       "      <td>12.01.2017</td>\n",
       "      <td>Minneapolis, MI</td>\n",
       "      <td>tcf_bank</td>\n",
       "      <td>1</td>\n",
       "      <td>Well I've been with TCF Bank for 3 plus years ...</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19270</th>\n",
       "      <td>edward</td>\n",
       "      <td>2.01.2017</td>\n",
       "      <td>Suite B, MI</td>\n",
       "      <td>tcf_bank</td>\n",
       "      <td>1</td>\n",
       "      <td>Deposited $800 3 days ago by certified check a...</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19271 rows × 7 columns</p>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-3a4524b1-8390-4758-af4b-6f84c694af69')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-3a4524b1-8390-4758-af4b-6f84c694af69 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-3a4524b1-8390-4758-af4b-6f84c694af69');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-ca06c021-e954-480b-9645-35e31be1d5cb\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ca06c021-e954-480b-9645-35e31be1d5cb')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-ca06c021-e954-480b-9645-35e31be1d5cb button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_aca372c9-70ab-4aa4-8fda-cc5b439100cf\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('data')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_aca372c9-70ab-4aa4-8fda-cc5b439100cf button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('data');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "        author        date         location          bank  star  \\\n",
       "0         Kyle  31.08.2023     Magnolia, TX  merrick_bank     5   \n",
       "1      Julicia  23.08.2023     Columbus, GA  merrick_bank     5   \n",
       "2        Karen   2.06.2023      Marrero, LA  merrick_bank     4   \n",
       "3        Brent  29.03.2023     Moultrie, GA  merrick_bank     5   \n",
       "4       Sharon  23.11.2022      Burnham, IL  merrick_bank     5   \n",
       "...        ...         ...              ...           ...   ...   \n",
       "19266       J.  30.01.2017        Salem, OR      tcf_bank     1   \n",
       "19267  Destiny  28.01.2017      Andover, MN      tcf_bank     1   \n",
       "19268     Sean  25.01.2017      Bothell, WA      tcf_bank     1   \n",
       "19269    Edgar  12.01.2017  Minneapolis, MI      tcf_bank     1   \n",
       "19270   edward   2.01.2017      Suite B, MI      tcf_bank     1   \n",
       "\n",
       "                                                    text  like  \n",
       "0      Very easy to use to view statements and make o...   NaN  \n",
       "1      Merrick Bank has always been good to me for bu...   NaN  \n",
       "2      Times are tough for everyone and I have worked...   3.0  \n",
       "3      I can not asked for a better Credit Card Compa...   3.0  \n",
       "4      Updated on 02/10/2023: I was happy to sign for...   3.0  \n",
       "...                                                  ...   ...  \n",
       "19266  Paid my 1st payment on time. They sent me a la...  11.0  \n",
       "19267  I have banked with TCF for about 4 years now a...  12.0  \n",
       "19268  Most inconvenient bank ever. As a business own...  10.0  \n",
       "19269  Well I've been with TCF Bank for 3 plus years ...  12.0  \n",
       "19270  Deposited $800 3 days ago by certified check a...  10.0  \n",
       "\n",
       "[19271 rows x 7 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Import the dataset\n",
    "data = pd.read_csv('/content/drive/My Drive/Colab Notebooks/Banks.csv')\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0M9Ck85s4nv4"
   },
   "source": [
    "## 2. Data Preprocessing\n",
    "\n",
    "- Add a sentiment column indicating positive and negative reviews based on a specified criteria (e.g., 4-5 stars as positive, 1-3 stars as negative).\n",
    "- Remove punctuations and numbers from the text.\n",
    "- Convert the text into lowercase for uniformity."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1701409890062,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "HHwd8dJ288-M",
    "outputId": "77ec986c-75ec-4a62-c7fd-7b587a2b8717"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "author         1\n",
       "date           0\n",
       "location       0\n",
       "bank           0\n",
       "star           0\n",
       "text          90\n",
       "like        1502\n",
       "dtype: int64"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Let's check for missing values\n",
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "executionInfo": {
     "elapsed": 5,
     "status": "ok",
     "timestamp": 1701409890062,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "cnOOlKAk9cBN"
   },
   "outputs": [],
   "source": [
    "# Delete rows that contain missing values\n",
    "data = data.dropna(subset=['text'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 10,
     "status": "ok",
     "timestamp": 1701409890709,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "0tqI0u3A9ngV",
    "outputId": "eb914b0c-6838-4cd2-a24b-5d61630cc12f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "author         1\n",
       "date           0\n",
       "location       0\n",
       "bank           0\n",
       "star           0\n",
       "text           0\n",
       "like        1496\n",
       "dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Recheck for missing values\n",
    "data.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 528
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1701409890710,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "y4Vq-U6q5AVn",
    "outputId": "d8c192a2-1f87-4de1-93da-949ca0d10a3c"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-7-f676d0e68b29>:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  reviews_data['sentiment'] = reviews_data['star'].apply(lambda x: 0 if x <= 3 else 1)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-0fd44a61-0d3e-48b6-9b61-aff191279510\" class=\"colab-df-container\">\n",
       "    <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>star</th>\n",
       "      <th>text</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>Very easy to use to view statements and make o...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>Merrick Bank has always been good to me for bu...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>Times are tough for everyone and I have worked...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>I can not asked for a better Credit Card Compa...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>Updated on 02/10/2023: I was happy to sign for...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19266</th>\n",
       "      <td>1</td>\n",
       "      <td>Paid my 1st payment on time. They sent me a la...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19267</th>\n",
       "      <td>1</td>\n",
       "      <td>I have banked with TCF for about 4 years now a...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19268</th>\n",
       "      <td>1</td>\n",
       "      <td>Most inconvenient bank ever. As a business own...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19269</th>\n",
       "      <td>1</td>\n",
       "      <td>Well I've been with TCF Bank for 3 plus years ...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19270</th>\n",
       "      <td>1</td>\n",
       "      <td>Deposited $800 3 days ago by certified check a...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19181 rows × 3 columns</p>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0fd44a61-0d3e-48b6-9b61-aff191279510')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-0fd44a61-0d3e-48b6-9b61-aff191279510 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-0fd44a61-0d3e-48b6-9b61-aff191279510');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-8b3f06ab-6c63-4e31-9fa3-ff1c39e3b6be\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-8b3f06ab-6c63-4e31-9fa3-ff1c39e3b6be')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-8b3f06ab-6c63-4e31-9fa3-ff1c39e3b6be button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_01c91985-b5f9-4556-b680-d411594d672f\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('reviews_data')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_01c91985-b5f9-4556-b680-d411594d672f button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('reviews_data');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "       star                                               text  sentiment\n",
       "0         5  Very easy to use to view statements and make o...          1\n",
       "1         5  Merrick Bank has always been good to me for bu...          1\n",
       "2         4  Times are tough for everyone and I have worked...          1\n",
       "3         5  I can not asked for a better Credit Card Compa...          1\n",
       "4         5  Updated on 02/10/2023: I was happy to sign for...          1\n",
       "...     ...                                                ...        ...\n",
       "19266     1  Paid my 1st payment on time. They sent me a la...          0\n",
       "19267     1  I have banked with TCF for about 4 years now a...          0\n",
       "19268     1  Most inconvenient bank ever. As a business own...          0\n",
       "19269     1  Well I've been with TCF Bank for 3 plus years ...          0\n",
       "19270     1  Deposited $800 3 days ago by certified check a...          0\n",
       "\n",
       "[19181 rows x 3 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Consider 4-5 star reviews as positive reviews (1) and 1-3 star reviews as negative reviews (0)\n",
    "reviews_data = data[['star', 'text']]\n",
    "reviews_data['sentiment'] = reviews_data['star'].apply(lambda x: 0 if x <= 3 else 1)\n",
    "reviews_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F2mR4frq7PWH"
   },
   "source": [
    "**Clean the text**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 528
    },
    "executionInfo": {
     "elapsed": 9170,
     "status": "ok",
     "timestamp": 1701409899874,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "favW60U24eZ9",
    "outputId": "eef71243-1897-4dd3-9040-505c1ed7142e"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-8-b06ce47722c8>:39: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  reviews_data['text'] = reviews_data['text'].apply(clean_text)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-c2f82509-d272-434e-ba24-86d701850447\" class=\"colab-df-container\">\n",
       "    <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>star</th>\n",
       "      <th>text</th>\n",
       "      <th>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>easy use view statements make online payments...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>merrick bank always good business rely merric...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>4</td>\n",
       "      <td>times tough everyone worked hard get credit r...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>asked better credit card company merrick bank...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>updated happy sign new credit card merrick co...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19266</th>\n",
       "      <td>1</td>\n",
       "      <td>paid st payment time sent late notice stating...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19267</th>\n",
       "      <td>1</td>\n",
       "      <td>banked tcf years first bank account ever past...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19268</th>\n",
       "      <td>1</td>\n",
       "      <td>inconvenient bank ever business owner many ba...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19269</th>\n",
       "      <td>1</td>\n",
       "      <td>well ive tcf bank plus years never issue ive ...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19270</th>\n",
       "      <td>1</td>\n",
       "      <td>deposited days ago certified check electronic...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19181 rows × 3 columns</p>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-c2f82509-d272-434e-ba24-86d701850447')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-c2f82509-d272-434e-ba24-86d701850447 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-c2f82509-d272-434e-ba24-86d701850447');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-4139dadb-301c-46cb-b00a-1a5f254c454d\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-4139dadb-301c-46cb-b00a-1a5f254c454d')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-4139dadb-301c-46cb-b00a-1a5f254c454d button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_48528d96-8030-4e4c-8efc-1d6d46c5665e\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('reviews_data')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_48528d96-8030-4e4c-8efc-1d6d46c5665e button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('reviews_data');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "       star                                               text  sentiment\n",
       "0         5   easy use view statements make online payments...          1\n",
       "1         5   merrick bank always good business rely merric...          1\n",
       "2         4   times tough everyone worked hard get credit r...          1\n",
       "3         5   asked better credit card company merrick bank...          1\n",
       "4         5   updated happy sign new credit card merrick co...          1\n",
       "...     ...                                                ...        ...\n",
       "19266     1   paid st payment time sent late notice stating...          0\n",
       "19267     1   banked tcf years first bank account ever past...          0\n",
       "19268     1   inconvenient bank ever business owner many ba...          0\n",
       "19269     1   well ive tcf bank plus years never issue ive ...          0\n",
       "19270     1   deposited days ago certified check electronic...          0\n",
       "\n",
       "[19181 rows x 3 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Function to clean the text\n",
    "def clean_text(text):\n",
    "\n",
    "    \"\"\" A function to clean the text \"\"\"\n",
    "\n",
    "    # Convert all text to lowercase to ensure uniformity\n",
    "    text = str(text).lower()\n",
    "\n",
    "    # Remove punctuations, and numbers\n",
    "    text = re.sub(r'[^a-zA-Z0-9\\s]', '', text)\n",
    "    text = re.sub(r'\\d', '', text)\n",
    "\n",
    "    # Create a set of stopwords\n",
    "    stop_words = set(stopwords.words('english'))\n",
    "\n",
    "    # Tokenize the text into words\n",
    "    words = tokenization(text)\n",
    "    cleaned_text = ''\n",
    "    for word in words:\n",
    "\n",
    "        # Get the token excludes those stopwords\n",
    "        if word not in stop_words:\n",
    "            cleaned_text += ' ' + word\n",
    "\n",
    "    return cleaned_text\n",
    "\n",
    "\n",
    "# Function to tokenize the text\n",
    "def tokenization(text):\n",
    "\n",
    "    \"\"\" A function to tokenize the text \"\"\"\n",
    "\n",
    "    words = nltk.word_tokenize(text)\n",
    "\n",
    "    return words\n",
    "\n",
    "\n",
    "# Call the function to clean and tokenize the text.\n",
    "reviews_data['text'] = reviews_data['text'].apply(clean_text)\n",
    "reviews_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Rl9p8luLBy5b"
   },
   "source": [
    "**Data statistics**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 125
    },
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1701409899875,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "HNfkiDEnALCi",
    "outputId": "bcf47122-8561-45e0-fb1d-8951eb2d3bd1"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-99755b7f-35d6-4b8f-a0ef-90d154960de1\" class=\"colab-df-container\">\n",
       "    <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>Sentiment</th>\n",
       "      <th>#Reviews</th>\n",
       "      <th>Min. #tokens</th>\n",
       "      <th>Avg. #tokens</th>\n",
       "      <th>Max. #tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Positive reviews</td>\n",
       "      <td>1869</td>\n",
       "      <td>8</td>\n",
       "      <td>212</td>\n",
       "      <td>643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Negative reviews</td>\n",
       "      <td>17312</td>\n",
       "      <td>6</td>\n",
       "      <td>349</td>\n",
       "      <td>1071</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-99755b7f-35d6-4b8f-a0ef-90d154960de1')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-99755b7f-35d6-4b8f-a0ef-90d154960de1 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-99755b7f-35d6-4b8f-a0ef-90d154960de1');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-7ce0efec-a610-4875-917c-fd645575b933\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-7ce0efec-a610-4875-917c-fd645575b933')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-7ce0efec-a610-4875-917c-fd645575b933 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_8a1ccb30-fdf2-4374-8272-6e565d381a1c\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('dataset_info_table')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_8a1ccb30-fdf2-4374-8272-6e565d381a1c button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('dataset_info_table');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "          Sentiment  #Reviews  Min. #tokens  Avg. #tokens  Max. #tokens\n",
       "0  Positive reviews      1869             8           212           643\n",
       "1  Negative reviews     17312             6           349          1071"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the number of tokens\n",
    "def get_num_tokens(tokens_list):\n",
    "\n",
    "    \"\"\"A function to get the number of tokens\"\"\"\n",
    "\n",
    "    return len(tokens_list)\n",
    "\n",
    "# Calculate basic statistics of the data\n",
    "dataset_info = {\n",
    "    'Sentiment': ['Positive reviews',\n",
    "                  'Negative reviews'],\n",
    "    '#Reviews': [len(reviews_data[reviews_data['sentiment'] == 1]['text']),\n",
    "                 len(reviews_data[reviews_data['sentiment'] == 0]['text'])],\n",
    "    'Min. #tokens': [min(reviews_data[reviews_data['sentiment'] == 1]['text'].apply(get_num_tokens)),\n",
    "                     min(reviews_data[reviews_data['sentiment'] == 0]['text'].apply(get_num_tokens))],\n",
    "    'Avg. #tokens': [round(reviews_data[reviews_data['sentiment'] == 1]['text'].apply(get_num_tokens).mean()),\n",
    "                     round(reviews_data[reviews_data['sentiment'] == 0]['text'].apply(get_num_tokens).mean())],\n",
    "    'Max. #tokens': [max(reviews_data[reviews_data['sentiment'] == 1]['text'].apply(get_num_tokens)),\n",
    "                     max(reviews_data[reviews_data['sentiment'] == 0]['text'].apply(get_num_tokens))]\n",
    "}\n",
    "\n",
    "# Create a DataFrame that contains the statistics of the data\n",
    "dataset_info_table = pd.DataFrame(dataset_info)\n",
    "\n",
    "# Print the statistics table\n",
    "dataset_info_table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 472
    },
    "executionInfo": {
     "elapsed": 524,
     "status": "ok",
     "timestamp": 1701409900393,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "DLsI1dAGHrj3",
    "outputId": "8266561d-2479-4bed-f4da-387d7a7a6d94"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the bar chart to see the distribution of the number of tokens\n",
    "sequence_lengths = [reviews_data['text'].apply(get_num_tokens)]\n",
    "plt.hist(sequence_lengths, bins=30)\n",
    "plt.xlabel('Number of Tokens')\n",
    "plt.ylabel('Frequency')\n",
    "plt.title('Distribution of Number of Tokens in Sequences')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "St89v63KxRWJ"
   },
   "source": [
    "## 3. Tokenization and Model Training\n",
    "Systematically divided into three sets: training (80%), validation (10%), and test (10%). Training set instructs the BERT model, enabling it to learn patterns and relationships within the data. Validation set is crucial for fine-tuning the model, experimenting with hyperparameter configurations. Test set is for evaluating the model performance."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Gw31_TJO7Y5U"
   },
   "source": [
    "**Split the datset into 3 parts: Training set, Valiation set, Test set**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1701409900393,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "huycYjR-0hea"
   },
   "outputs": [],
   "source": [
    "# Define input variable and target variable\n",
    "reviews = pd.DataFrame(reviews_data['text'])\n",
    "labels = pd.DataFrame(reviews_data['sentiment'])\n",
    "\n",
    "# Split the dataset into 3 parts, including training set (80%), validation set (10%), and test set (10%)\n",
    "reviews_train, reviews_temp, labels_train, labels_temp = train_test_split(reviews, labels, test_size=0.2, random_state=100)\n",
    "reviews_val, reviews_test, labels_val, labels_test = train_test_split(reviews_temp, labels_temp, test_size=0.5, random_state=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 424
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1701409900393,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "eYA-tVPmLA6M",
    "outputId": "b05f7d45-3491-4d6b-89e6-3d77ac7d50db"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-eef53cae-479e-4e74-b613-29bc660e9c8e\" class=\"colab-df-container\">\n",
       "    <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>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>easy use view statements make online payments...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>merrick bank always good business rely merric...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>times tough everyone worked hard get credit r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>asked better credit card company merrick bank...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>updated happy sign new credit card merrick co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19266</th>\n",
       "      <td>paid st payment time sent late notice stating...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19267</th>\n",
       "      <td>banked tcf years first bank account ever past...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19268</th>\n",
       "      <td>inconvenient bank ever business owner many ba...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19269</th>\n",
       "      <td>well ive tcf bank plus years never issue ive ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19270</th>\n",
       "      <td>deposited days ago certified check electronic...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19181 rows × 1 columns</p>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-eef53cae-479e-4e74-b613-29bc660e9c8e')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-eef53cae-479e-4e74-b613-29bc660e9c8e button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-eef53cae-479e-4e74-b613-29bc660e9c8e');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-13dd7a8d-1124-4af7-81bd-e5e8d5958894\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-13dd7a8d-1124-4af7-81bd-e5e8d5958894')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-13dd7a8d-1124-4af7-81bd-e5e8d5958894 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_bc39702e-7522-4a7a-ac98-74ecf48d9b92\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('reviews')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_bc39702e-7522-4a7a-ac98-74ecf48d9b92 button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('reviews');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "                                                    text\n",
       "0       easy use view statements make online payments...\n",
       "1       merrick bank always good business rely merric...\n",
       "2       times tough everyone worked hard get credit r...\n",
       "3       asked better credit card company merrick bank...\n",
       "4       updated happy sign new credit card merrick co...\n",
       "...                                                  ...\n",
       "19266   paid st payment time sent late notice stating...\n",
       "19267   banked tcf years first bank account ever past...\n",
       "19268   inconvenient bank ever business owner many ba...\n",
       "19269   well ive tcf bank plus years never issue ive ...\n",
       "19270   deposited days ago certified check electronic...\n",
       "\n",
       "[19181 rows x 1 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reviews"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 424
    },
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1701409900393,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "XHddFf52LCk_",
    "outputId": "d0a86038-3edb-4c17-81b6-84ec04bc479b"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-22c05727-8eb3-4885-9fa7-50ff0efb0691\" class=\"colab-df-container\">\n",
       "    <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>sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19266</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19267</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19268</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19269</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19270</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19181 rows × 1 columns</p>\n",
       "</div>\n",
       "    <div class=\"colab-df-buttons\">\n",
       "\n",
       "  <div class=\"colab-df-container\">\n",
       "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-22c05727-8eb3-4885-9fa7-50ff0efb0691')\"\n",
       "            title=\"Convert this dataframe to an interactive table.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
       "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "\n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    .colab-df-buttons div {\n",
       "      margin-bottom: 4px;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "    <script>\n",
       "      const buttonEl =\n",
       "        document.querySelector('#df-22c05727-8eb3-4885-9fa7-50ff0efb0691 button.colab-df-convert');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      async function convertToInteractive(key) {\n",
       "        const element = document.querySelector('#df-22c05727-8eb3-4885-9fa7-50ff0efb0691');\n",
       "        const dataTable =\n",
       "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                    [key], {});\n",
       "        if (!dataTable) return;\n",
       "\n",
       "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "          + ' to learn more about interactive tables.';\n",
       "        element.innerHTML = '';\n",
       "        dataTable['output_type'] = 'display_data';\n",
       "        await google.colab.output.renderOutput(dataTable, element);\n",
       "        const docLink = document.createElement('div');\n",
       "        docLink.innerHTML = docLinkHtml;\n",
       "        element.appendChild(docLink);\n",
       "      }\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "\n",
       "<div id=\"df-1a6b349e-06ff-4ec5-8f7e-ea3063b42ad4\">\n",
       "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-1a6b349e-06ff-4ec5-8f7e-ea3063b42ad4')\"\n",
       "            title=\"Suggest charts\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "     width=\"24px\">\n",
       "    <g>\n",
       "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
       "    </g>\n",
       "</svg>\n",
       "  </button>\n",
       "\n",
       "<style>\n",
       "  .colab-df-quickchart {\n",
       "      --bg-color: #E8F0FE;\n",
       "      --fill-color: #1967D2;\n",
       "      --hover-bg-color: #E2EBFA;\n",
       "      --hover-fill-color: #174EA6;\n",
       "      --disabled-fill-color: #AAA;\n",
       "      --disabled-bg-color: #DDD;\n",
       "  }\n",
       "\n",
       "  [theme=dark] .colab-df-quickchart {\n",
       "      --bg-color: #3B4455;\n",
       "      --fill-color: #D2E3FC;\n",
       "      --hover-bg-color: #434B5C;\n",
       "      --hover-fill-color: #FFFFFF;\n",
       "      --disabled-bg-color: #3B4455;\n",
       "      --disabled-fill-color: #666;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart {\n",
       "    background-color: var(--bg-color);\n",
       "    border: none;\n",
       "    border-radius: 50%;\n",
       "    cursor: pointer;\n",
       "    display: none;\n",
       "    fill: var(--fill-color);\n",
       "    height: 32px;\n",
       "    padding: 0;\n",
       "    width: 32px;\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart:hover {\n",
       "    background-color: var(--hover-bg-color);\n",
       "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "    fill: var(--button-hover-fill-color);\n",
       "  }\n",
       "\n",
       "  .colab-df-quickchart-complete:disabled,\n",
       "  .colab-df-quickchart-complete:disabled:hover {\n",
       "    background-color: var(--disabled-bg-color);\n",
       "    fill: var(--disabled-fill-color);\n",
       "    box-shadow: none;\n",
       "  }\n",
       "\n",
       "  .colab-df-spinner {\n",
       "    border: 2px solid var(--fill-color);\n",
       "    border-color: transparent;\n",
       "    border-bottom-color: var(--fill-color);\n",
       "    animation:\n",
       "      spin 1s steps(1) infinite;\n",
       "  }\n",
       "\n",
       "  @keyframes spin {\n",
       "    0% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "      border-left-color: var(--fill-color);\n",
       "    }\n",
       "    20% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    30% {\n",
       "      border-color: transparent;\n",
       "      border-left-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    40% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-top-color: var(--fill-color);\n",
       "    }\n",
       "    60% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "    }\n",
       "    80% {\n",
       "      border-color: transparent;\n",
       "      border-right-color: var(--fill-color);\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "    90% {\n",
       "      border-color: transparent;\n",
       "      border-bottom-color: var(--fill-color);\n",
       "    }\n",
       "  }\n",
       "</style>\n",
       "\n",
       "  <script>\n",
       "    async function quickchart(key) {\n",
       "      const quickchartButtonEl =\n",
       "        document.querySelector('#' + key + ' button');\n",
       "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
       "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
       "      try {\n",
       "        const charts = await google.colab.kernel.invokeFunction(\n",
       "            'suggestCharts', [key], {});\n",
       "      } catch (error) {\n",
       "        console.error('Error during call to suggestCharts:', error);\n",
       "      }\n",
       "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
       "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
       "    }\n",
       "    (() => {\n",
       "      let quickchartButtonEl =\n",
       "        document.querySelector('#df-1a6b349e-06ff-4ec5-8f7e-ea3063b42ad4 button');\n",
       "      quickchartButtonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "    })();\n",
       "  </script>\n",
       "</div>\n",
       "\n",
       "  <div id=\"id_484e1f02-aabe-4011-a42a-2718d7a6de80\">\n",
       "    <style>\n",
       "      .colab-df-generate {\n",
       "        background-color: #E8F0FE;\n",
       "        border: none;\n",
       "        border-radius: 50%;\n",
       "        cursor: pointer;\n",
       "        display: none;\n",
       "        fill: #1967D2;\n",
       "        height: 32px;\n",
       "        padding: 0 0 0 0;\n",
       "        width: 32px;\n",
       "      }\n",
       "\n",
       "      .colab-df-generate:hover {\n",
       "        background-color: #E2EBFA;\n",
       "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "        fill: #174EA6;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate {\n",
       "        background-color: #3B4455;\n",
       "        fill: #D2E3FC;\n",
       "      }\n",
       "\n",
       "      [theme=dark] .colab-df-generate:hover {\n",
       "        background-color: #434B5C;\n",
       "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "        fill: #FFFFFF;\n",
       "      }\n",
       "    </style>\n",
       "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('labels')\"\n",
       "            title=\"Generate code using this dataframe.\"\n",
       "            style=\"display:none;\">\n",
       "\n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
       "  </svg>\n",
       "    </button>\n",
       "    <script>\n",
       "      (() => {\n",
       "      const buttonEl =\n",
       "        document.querySelector('#id_484e1f02-aabe-4011-a42a-2718d7a6de80 button.colab-df-generate');\n",
       "      buttonEl.style.display =\n",
       "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "      buttonEl.onclick = () => {\n",
       "        google.colab.notebook.generateWithVariable('labels');\n",
       "      }\n",
       "      })();\n",
       "    </script>\n",
       "  </div>\n",
       "\n",
       "    </div>\n",
       "  </div>\n"
      ],
      "text/plain": [
       "       sentiment\n",
       "0              1\n",
       "1              1\n",
       "2              1\n",
       "3              1\n",
       "4              1\n",
       "...          ...\n",
       "19266          0\n",
       "19267          0\n",
       "19268          0\n",
       "19269          0\n",
       "19270          0\n",
       "\n",
       "[19181 rows x 1 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F7TKg0It7fh8"
   },
   "source": [
    "**Tokenize and encode the data using the BERT tokenizer**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "executionInfo": {
     "elapsed": 24204,
     "status": "ok",
     "timestamp": 1701409924590,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "i_UFQ4cx1tmk"
   },
   "outputs": [],
   "source": [
    "# Load pre-trained BERT tokenizer\n",
    "tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')\n",
    "\n",
    "# Tokenize and encode the sentences for the training set, validation set, and test set\n",
    "max_len = 256\n",
    "train_encodings = tokenizer(reviews_train['text'].tolist(), padding=True, truncation=True, return_tensors='pt')\n",
    "val_encodings = tokenizer(reviews_val['text'].tolist(), padding=True, truncation=True, return_tensors='pt')\n",
    "test_encodings = tokenizer(reviews_test['text'].tolist(), padding=True, truncation=True, return_tensors='pt')\n",
    "\n",
    "# Convert sentiment labels to PyTorch tensors\n",
    "train_labels = torch.tensor(labels_train['sentiment'].tolist())\n",
    "val_labels = torch.tensor(labels_val['sentiment'].tolist())\n",
    "test_labels = torch.tensor(labels_test['sentiment'].tolist())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 25,
     "status": "ok",
     "timestamp": 1701409924590,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "_fusrRgQE1pt",
    "outputId": "5ff8630a-fd69-4855-98ff-9873cecf2c30"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'input_ids': tensor([[  101,  5223, 12978,  ...,     0,     0,     0],\n",
       "        [  101,  2034,  5741,  ...,     0,     0,     0],\n",
       "        [  101,  1052, 12273,  ...,     0,     0,     0],\n",
       "        ...,\n",
       "        [  101,  2288,  4003,  ...,     0,     0,     0],\n",
       "        [  101,  4162,  3784,  ...,     0,     0,     0],\n",
       "        [  101, 10263,  3361,  ...,     0,     0,     0]]), 'token_type_ids': tensor([[0, 0, 0,  ..., 0, 0, 0],\n",
       "        [0, 0, 0,  ..., 0, 0, 0],\n",
       "        [0, 0, 0,  ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0,  ..., 0, 0, 0],\n",
       "        [0, 0, 0,  ..., 0, 0, 0],\n",
       "        [0, 0, 0,  ..., 0, 0, 0]]), 'attention_mask': tensor([[1, 1, 1,  ..., 0, 0, 0],\n",
       "        [1, 1, 1,  ..., 0, 0, 0],\n",
       "        [1, 1, 1,  ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [1, 1, 1,  ..., 0, 0, 0],\n",
       "        [1, 1, 1,  ..., 0, 0, 0],\n",
       "        [1, 1, 1,  ..., 0, 0, 0]])}"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_encodings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OEcrRFbY7kF-"
   },
   "source": [
    "**Create DataLoader**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "executionInfo": {
     "elapsed": 24,
     "status": "ok",
     "timestamp": 1701409924591,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "MQQTnF6qx3L5"
   },
   "outputs": [],
   "source": [
    "# Define datasets using TensorDataset, containing input_ids, attention_mask, and labels\n",
    "train_dataset = TensorDataset(train_encodings['input_ids'], train_encodings['attention_mask'], train_labels)\n",
    "val_dataset = TensorDataset(val_encodings['input_ids'], val_encodings['attention_mask'], val_labels)\n",
    "test_dataset = TensorDataset(test_encodings['input_ids'], test_encodings['attention_mask'], test_labels)\n",
    "\n",
    "# Define batch size\n",
    "batch_size = 32\n",
    "\n",
    "# Create DataLoader for training set , validation set, and test set\n",
    "train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)\n",
    "val_loader = DataLoader(val_dataset, batch_size=batch_size, shuffle=False)\n",
    "test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 970,
     "status": "ok",
     "timestamp": 1701409925538,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "oeg7OJvcynep",
    "outputId": "50573981-93a1-4a15-e40a-fca3f1da33f3"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of BertForSequenceClassification were not initialized from the model checkpoint at bert-base-uncased and are newly initialized: ['classifier.weight', 'classifier.bias']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
     ]
    }
   ],
   "source": [
    "# Load Pre-trained BERT Model\n",
    "model = BertForSequenceClassification.from_pretrained('bert-base-uncased', num_labels=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MButFS9q6wxN"
   },
   "source": [
    "**Fine-Tuning BERT model using Validation Set**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 874
    },
    "executionInfo": {
     "elapsed": 403741,
     "status": "ok",
     "timestamp": 1701410329276,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "pE-I0sNXy6jP",
    "outputId": "b3ce8a0b-8067-4fcf-ce8d-d8809eb4e487"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/transformers/optimization.py:411: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use the PyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/3, Training Loss: 0.15597056420519947, Validation Accuracy: 0.9718456725755996\n",
      "Training Time for Epoch 1: 129.44 seconds\n",
      "Max GPU Memory Allocated: 5745.69 MB\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.10/dist-packages/torch/cuda/memory.py:329: FutureWarning: torch.cuda.reset_max_memory_allocated now calls torch.cuda.reset_peak_memory_stats, which resets /all/ peak memory stats.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2/3, Training Loss: 0.07650929606170394, Validation Accuracy: 0.9718456725755996\n",
      "Training Time for Epoch 2: 129.41 seconds\n",
      "Max GPU Memory Allocated: 5748.17 MB\n",
      "Epoch 3/3, Training Loss: 0.0450174947368699, Validation Accuracy: 0.9708029197080292\n",
      "Training Time for Epoch 3: 129.37 seconds\n",
      "Max GPU Memory Allocated: 5744.78 MB\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+-------+---------------------+---------------------+--------------------+-------------------------------+\n",
      "| Epoch |    Training Loss    | Validation Accuracy | Training Time (s)  | Max GPU Memory Allocated (MB) |\n",
      "+-------+---------------------+---------------------+--------------------+-------------------------------+\n",
      "|   1   | 0.15597056420519947 |  0.9718456725755996 | 129.43725109100342 |           5745.6875           |\n",
      "|   2   | 0.07650929606170394 |  0.9718456725755996 | 129.4111795425415  |        5748.1650390625        |\n",
      "|   3   |  0.0450174947368699 |  0.9708029197080292 | 129.37072157859802 |        5744.7822265625        |\n",
      "+-------+---------------------+---------------------+--------------------+-------------------------------+\n"
     ]
    }
   ],
   "source": [
    "# Move the model to the appropriate device (GPU if available, otherwise CPU)\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "model.to(device)\n",
    "\n",
    "# Define the optimizer (AdamW) with a learning rate of 2e-5\n",
    "optimizer = AdamW(model.parameters(), lr=2e-5)\n",
    "\n",
    "# Lists to store values for plotting\n",
    "train_loss_values = []\n",
    "val_accuracy_values = []\n",
    "training_times = []\n",
    "memory_allocated_values = []\n",
    "\n",
    "# Initialize a PrettyTable\n",
    "table = PrettyTable()\n",
    "table.field_names = [\"Epoch\", \"Training Loss\", \"Validation Accuracy\", \"Training Time (s)\", \"Max GPU Memory Allocated (MB)\"]\n",
    "\n",
    "# Number of training epochs\n",
    "num_epochs = 3\n",
    "\n",
    "# Loop through each epoch\n",
    "for epoch in range(num_epochs):\n",
    "\n",
    "    # Record the start time\n",
    "    start_time = time.time()\n",
    "\n",
    "    # Set the model to training mode and initialize total loss for the epoch\n",
    "    model.train()\n",
    "    total_loss = 0.0\n",
    "\n",
    "    # Loop through the training data in batches\n",
    "    for batch in train_loader:\n",
    "\n",
    "        # Move input data to the appropriate device\n",
    "        input_ids, attention_mask, labels = batch\n",
    "        input_ids, attention_mask, labels = input_ids.to(device), attention_mask.to(device), labels.to(device)\n",
    "\n",
    "        # Zero the gradients to prevent accumulation from previous batches\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        # Forward pass: calculate model predictions and compute the loss\n",
    "        outputs = model(input_ids=input_ids, attention_mask=attention_mask, labels=labels)\n",
    "        loss = F.cross_entropy(outputs.logits, labels)\n",
    "        total_loss += loss.item()\n",
    "\n",
    "        # Backward pass: compute gradient of the loss with respect to model parameters\n",
    "        loss.backward()\n",
    "\n",
    "        # Optimization step: update model parameters using the optimizer\n",
    "        optimizer.step()\n",
    "\n",
    "    # Calculate average loss for the epoch\n",
    "    average_loss = total_loss / len(train_loader)\n",
    "\n",
    "    # Record the end time\n",
    "    end_time = time.time()\n",
    "\n",
    "    # Calculate training time\n",
    "    training_time = end_time - start_time\n",
    "\n",
    "    # Set the model to evaluation mode to evaluate the performance of each epoch\n",
    "    model.eval()\n",
    "\n",
    "    # Initialize total correct predictions and total samples for validation\n",
    "    total_correct = 0\n",
    "    total_samples = 0\n",
    "\n",
    "    # Disable gradient computation during validation\n",
    "    with torch.no_grad():\n",
    "\n",
    "        # Loop through the validation data in batches\n",
    "        for batch in val_loader:\n",
    "\n",
    "            # Move input data to the appropriate device\n",
    "            input_ids, attention_mask, labels = batch\n",
    "            input_ids, attention_mask, labels = input_ids.to(device), attention_mask.to(device), labels.to(device)\n",
    "\n",
    "            # Forward pass: calculate model predictions\n",
    "            outputs = model(input_ids, attention_mask=attention_mask)\n",
    "            logits = outputs.logits\n",
    "            predictions = torch.argmax(logits, dim=1)\n",
    "\n",
    "            # Update total correct predictions and total samples\n",
    "            total_correct += torch.sum(predictions == labels).item()\n",
    "            total_samples += labels.size(0)\n",
    "\n",
    "    # Calculate validation accuracy\n",
    "    accuracy = total_correct / total_samples\n",
    "\n",
    "    # Record the end time\n",
    "    end_time = time.time()\n",
    "\n",
    "    # Check GPU memory (MB)\n",
    "    max_memory_allocated = torch.cuda.max_memory_allocated() / 1024 ** 2\n",
    "\n",
    "    # Print result of the epoch\n",
    "    print(f'Epoch {epoch + 1}/{num_epochs}, Training Loss: {average_loss}, Validation Accuracy: {accuracy}')\n",
    "    print(f'Training Time for Epoch {epoch + 1}: {training_time:.2f} seconds')\n",
    "    print(f'Max GPU Memory Allocated: {max_memory_allocated:.2f} MB')\n",
    "\n",
    "    # Add values to the lists for plotting\n",
    "    train_loss_values.append(average_loss)\n",
    "    val_accuracy_values.append(accuracy)\n",
    "    training_times.append(training_time)\n",
    "    memory_allocated_values.append(max_memory_allocated)\n",
    "\n",
    "    # Add the results to the PrettyTable\n",
    "    table.add_row([epoch + 1, average_loss, accuracy, training_time, max_memory_allocated])\n",
    "\n",
    "    # Reser GPU memory statistics\n",
    "    torch.cuda.reset_max_memory_allocated()\n",
    "\n",
    "# Plotting the graphs\n",
    "epochs_range = range(1, num_epochs + 1)\n",
    "\n",
    "# Set the size of the graph\n",
    "plt.figure(figsize=(12, 5))\n",
    "\n",
    "# Plot Training Loss\n",
    "plt.subplot(1, 2, 1)\n",
    "plt.plot(epochs_range, train_loss_values, label='Training Loss', marker='o')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Training Loss')\n",
    "plt.title('Training Loss Over Epochs')\n",
    "plt.legend()\n",
    "\n",
    "# Plot Validation Accuracy\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(epochs_range, val_accuracy_values, label='Validation Accuracy', marker='o', color='orange')\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Validation Accuracy')\n",
    "plt.title('Validation Accuracy Over Epochs')\n",
    "plt.legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()\n",
    "\n",
    "# Print the table\n",
    "print(table)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "OYMh_x1U7rlr"
   },
   "source": [
    "## 4. Classification and Evaluation\n",
    "Evaluate the model performance to classify the reviews between positive and negative reviews on Test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 753
    },
    "executionInfo": {
     "elapsed": 3795,
     "status": "ok",
     "timestamp": 1701410333066,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "n02inld31qea",
    "outputId": "5424704f-38c0-46f4-c3f2-73370d27fc6f"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ROC-AUC Score: 0.9601\n",
      "Accuracy: 0.9620\n"
     ]
    }
   ],
   "source": [
    "# Set the model to evaluation mode\n",
    "model.eval()\n",
    "\n",
    "# Lists to store true labels and predicted probabilities\n",
    "true_labels = []\n",
    "predicted_probs = []\n",
    "predicted_labels = []\n",
    "\n",
    "# Evaluate the Model on the Test Set\n",
    "with torch.no_grad():\n",
    "    for batch in test_loader:\n",
    "        input_ids, attention_mask, labels = batch\n",
    "        input_ids, attention_mask, labels = input_ids.to(device), attention_mask.to(device), labels.to(device)\n",
    "\n",
    "        # Forward pass\n",
    "        outputs = model(input_ids, attention_mask=attention_mask)\n",
    "        logits = outputs.logits\n",
    "        probabilities = torch.softmax(logits, dim=1)[:, 1]\n",
    "        predictions = torch.argmax(logits, dim=1)\n",
    "\n",
    "        # Collect true labels and predicted probabilities\n",
    "        true_labels.extend(labels.cpu().numpy())\n",
    "        predicted_probs.extend(probabilities.cpu().numpy())\n",
    "        predicted_labels.extend(predictions.cpu().numpy())\n",
    "\n",
    "# Calculate ROC curve\n",
    "fpr, tpr, thresholds = roc_curve(true_labels, predicted_probs)\n",
    "\n",
    "# Calculate ROC-AUC score\n",
    "roc_auc = roc_auc_score(true_labels, predicted_probs)\n",
    "\n",
    "# Calculate accuracy\n",
    "accuracy = accuracy_score(true_labels, predicted_labels)\n",
    "\n",
    "# Plot ROC curve\n",
    "plt.figure(figsize=(8, 8))\n",
    "plt.plot(fpr, tpr, color='darkorange', lw=2, label=f'ROC Curve (AUC = {roc_auc:.2f})')\n",
    "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver Operating Characteristic (ROC) Curve')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.show()\n",
    "\n",
    "# Print ROC-AUC Score and Accuracy\n",
    "print(f'ROC-AUC Score: {roc_auc:.4f}')\n",
    "print(f'Accuracy: {accuracy:.4f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DrInRouY8i-_"
   },
   "source": [
    "**Manual evaluation**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 29,
     "status": "ok",
     "timestamp": 1701410333067,
     "user": {
      "displayName": "Thanapoom Phatthanaphan",
      "userId": "06703220080806313983"
     },
     "user_tz": 300
    },
    "id": "PV_eTLhw8lrw",
    "outputId": "df2f4cc0-0db4-4e2a-8624-36731c218e1e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Review: Impressed with the bank's online services, making transactions has never been easier.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: Customer support was unhelpful and frustrating; it took forever to resolve a simple issue.\n",
      "Predicted Sentiment: Negative\n",
      "True Sentiment: Negative\n",
      "\n",
      "Review: Received a great interest rate on my savings account, very satisfied with the banking experience.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: The mobile app is user-friendly and convenient for managing accounts on the go.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: Unexpected fees and hidden charges made my experience with this bank disappointing.\n",
      "Predicted Sentiment: Negative\n",
      "True Sentiment: Negative\n",
      "\n",
      "Review: Quick and efficient loan approval process, would recommend for financial assistance.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: The staff at the local branch were friendly and assisted me with professionalism.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: I've been a customer for years, and the bank has consistently met my financial needs.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: The credit card application process was straightforward, and I got approved quickly.\n",
      "Predicted Sentiment: Positive\n",
      "True Sentiment: Positive\n",
      "\n",
      "Review: Poor security measures; my account was compromised, and the recovery process was frustrating.\n",
      "Predicted Sentiment: Negative\n",
      "True Sentiment: Negative\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# 10 unseen reviews for manual test\n",
    "unseen_reviews = [\"Impressed with the bank's online services, making transactions has never been easier.\",\n",
    "                  \"Customer support was unhelpful and frustrating; it took forever to resolve a simple issue.\",\n",
    "                  \"Received a great interest rate on my savings account, very satisfied with the banking experience.\",\n",
    "                  \"The mobile app is user-friendly and convenient for managing accounts on the go.\",\n",
    "                  \"Unexpected fees and hidden charges made my experience with this bank disappointing.\",\n",
    "                  \"Quick and efficient loan approval process, would recommend for financial assistance.\",\n",
    "                  \"The staff at the local branch were friendly and assisted me with professionalism.\",\n",
    "                  \"I've been a customer for years, and the bank has consistently met my financial needs.\",\n",
    "                  \"The credit card application process was straightforward, and I got approved quickly.\",\n",
    "                  \"Poor security measures; my account was compromised, and the recovery process was frustrating.\"]\n",
    "\n",
    "# Sentiment for 10 unseen reviews\n",
    "sentiment_unseen_reviews = ['Positive', 'Negative', 'Positive', 'Positive', 'Negative', 'Positive', 'Positive', 'Positive', 'Positive', 'Negative']\n",
    "\n",
    "# Clean the text\n",
    "cleaned_unseen_reviews = []\n",
    "for text in unseen_reviews:\n",
    "    cleaned_text = clean_text(text)\n",
    "    cleaned_unseen_reviews.append(cleaned_text)\n",
    "\n",
    "# Tokenize and encode the unseen reviews\n",
    "unseen_encodings = tokenizer(unseen_reviews, padding=True, truncation=True, return_tensors='pt')\n",
    "\n",
    "# Make predictions\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    inputs = {key: unseen_encodings[key].to(model.device) for key in unseen_encodings}\n",
    "    outputs = model(**inputs)\n",
    "    logits = outputs.logits\n",
    "    predictions = torch.argmax(logits, dim=1)\n",
    "\n",
    "# Print the predicted sentiment for each review\n",
    "i = 0\n",
    "for review, prediction in zip(unseen_reviews, predictions.tolist()):\n",
    "    sentiment = \"Positive\" if prediction == 1 else \"Negative\"\n",
    "    print(f\"Review: {review}\\nPredicted Sentiment: {sentiment}\\nTrue Sentiment: {sentiment_unseen_reviews[i]}\\n\")\n",
    "    i += 1"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "gpuType": "V100",
   "machine_shape": "hm",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}