datacreation.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_kMfZv_2FGjO"
      },
      "source": [
        "# 📦 Notebook 1 — Real-World Data Extraction & Synthetic Enrichment\n",
        "### Project: E-Commerce Customer Intelligence — Food & Grocery Products\n",
        "**Business Question:** *How can an online food retailer reduce customer churn and optimize product positioning by analyzing review sentiment (qualitative) and rating/helpfulness patterns (quantitative)?*\n",
        "\n",
        "**Dataset:** Amazon Fine Food Reviews (Kaggle) — 568,000+ real customer reviews\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XHzEidhsFGjP"
      },
      "source": [
        "---\n",
        "## 📥 Step 0 — How to get the Kaggle dataset\n",
        "1. Go to: **https://www.kaggle.com/datasets/snap/amazon-fine-food-reviews**\n",
        "2. Click **Download** (free Kaggle account required)\n",
        "3. Extract the ZIP → you get **`Reviews.csv`**\n",
        "4. Upload `Reviews.csv` to Colab (folder icon → upload)\n",
        "5. Leave `CSV_FILE_PATH = \"Reviews.csv\"` as is\n",
        "\n",
        "> 💡 No Kaggle account? The built-in food sample loads automatically as fallback."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "v_7KefJ2FGjQ"
      },
      "source": [
        "---\n",
        "## ⚙️ Step 1 — Install & Import Libraries"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fp_U4-syFGjQ",
        "outputId": "24676d4a-9a89-416d-d213-a355fab148ed"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 2
        }
      ],
      "source": [
        "!pip install pandas matplotlib seaborn textblob -q\n",
        "import nltk\n",
        "nltk.download('punkt', quiet=True)\n",
        "nltk.download('averaged_perceptron_tagger', quiet=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SdBlUGWZFGjQ",
        "outputId": "ac8f917f-3337-4b60-971e-61a26da480ef"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ All libraries imported successfully\n"
          ]
        }
      ],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import random\n",
        "import os\n",
        "from textblob import TextBlob\n",
        "\n",
        "pd.set_option('display.max_colwidth', 120)\n",
        "plt.style.use('seaborn-v0_8-whitegrid')\n",
        "random.seed(42)\n",
        "np.random.seed(42)\n",
        "\n",
        "print('✅ All libraries imported successfully')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1xsy8-atFGjR"
      },
      "source": [
        "---\n",
        "## 🔑 Step 2 — Configuration"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Fwdd5VqeFGjR",
        "outputId": "6eae4756-fe27-4a6b-833e-6f191c7359b7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Config set — will sample 500 reviews for analysis\n"
          ]
        }
      ],
      "source": [
        "CSV_FILE_PATH   = \"Reviews.csv\"   # Kaggle Amazon Fine Food Reviews\n",
        "N_SAMPLE        = 500             # reviews to keep for analysis (adjust up to 2000 if you want more)\n",
        "RANDOM_SEED     = 42\n",
        "\n",
        "print(f'✅ Config set — will sample {N_SAMPLE} reviews for analysis')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xl06KX7IFGjR"
      },
      "source": [
        "---\n",
        "## 📂 Step 3 — Load Real-World Data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 703
        },
        "id": "BtNjKFlgFGjR",
        "outputId": "61efa873-1899-49ef-d972-f06794ee35e0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "⚠️  No CSV found — using built-in food reviews sample...\n",
            "✅ Built-in food sample loaded: 500 rows\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  ProductId            product_name       category  Score  \\\n",
              "0      F001       Organic Green Tea      Beverages      5   \n",
              "1      F001       Organic Green Tea      Beverages      1   \n",
              "2      F002  Dark Chocolate Bar 70%  Confectionery      5   \n",
              "\n",
              "                        Summary  \\\n",
              "0      Best tea I have ever had   \n",
              "1   Arrived stale and tasteless   \n",
              "2  Incredible quality chocolate   \n",
              "\n",
              "                                                                                                                      Text  \\\n",
              "0  This organic green tea is absolutely wonderful. The flavour is delicate and fresh, nothing like the bitter supermark...   \n",
              "1  Extremely disappointed. The tea had no aroma whatsoever and tasted like hot water. The bags were clearly old stock. ...   \n",
              "2  This is hands down the best dark chocolate I have tasted outside of a specialist chocolatier. The bitterness is perf...   \n",
              "\n",
              "   HelpfulnessNumerator  HelpfulnessDenominator  \n",
              "0                    35                      36  \n",
              "1                    26                      30  \n",
              "2                    59                      54  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-991b4960-ccf5-4530-9b71-a43406d041a3\" 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>ProductId</th>\n",
              "      <th>product_name</th>\n",
              "      <th>category</th>\n",
              "      <th>Score</th>\n",
              "      <th>Summary</th>\n",
              "      <th>Text</th>\n",
              "      <th>HelpfulnessNumerator</th>\n",
              "      <th>HelpfulnessDenominator</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>F001</td>\n",
              "      <td>Organic Green Tea</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Best tea I have ever had</td>\n",
              "      <td>This organic green tea is absolutely wonderful. The flavour is delicate and fresh, nothing like the bitter supermark...</td>\n",
              "      <td>35</td>\n",
              "      <td>36</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>F001</td>\n",
              "      <td>Organic Green Tea</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>1</td>\n",
              "      <td>Arrived stale and tasteless</td>\n",
              "      <td>Extremely disappointed. The tea had no aroma whatsoever and tasted like hot water. The bags were clearly old stock. ...</td>\n",
              "      <td>26</td>\n",
              "      <td>30</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>F002</td>\n",
              "      <td>Dark Chocolate Bar 70%</td>\n",
              "      <td>Confectionery</td>\n",
              "      <td>5</td>\n",
              "      <td>Incredible quality chocolate</td>\n",
              "      <td>This is hands down the best dark chocolate I have tasted outside of a specialist chocolatier. The bitterness is perf...</td>\n",
              "      <td>59</td>\n",
              "      <td>54</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-991b4960-ccf5-4530-9b71-a43406d041a3')\"\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-991b4960-ccf5-4530-9b71-a43406d041a3 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-991b4960-ccf5-4530-9b71-a43406d041a3');\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>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df_raw",
              "summary": "{\n  \"name\": \"df_raw\",\n  \"rows\": 500,\n  \"fields\": [\n    {\n      \"column\": \"ProductId\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"F009\",\n          \"F002\",\n          \"F006\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"product_name\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 10,\n        \"samples\": [\n          \"Protein Granola\",\n          \"Dark Chocolate Bar 70%\",\n          \"Instant Ramen Variety Pack\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"category\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 8,\n        \"samples\": [\n          \"Confectionery\",\n          \"Noodles\",\n          \"Beverages\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Score\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 1,\n        \"max\": 5,\n        \"num_unique_values\": 5,\n        \"samples\": [\n          1,\n          4,\n          2\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Summary\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 20,\n        \"samples\": [\n          \"Best tea I have ever had\",\n          \"Artificial sweetener aftertaste ruins it\",\n          \"Oil separation is extreme and annoying\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Text\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 20,\n        \"samples\": [\n          \"This organic green tea is absolutely wonderful. The flavour is delicate and fresh, nothing like the bitter supermarket brands. I have been drinking it every morning for 3 months and will never switch.\",\n          \"The macros look great on paper but the stevia aftertaste is so strong it makes the whole bowl unpleasant. I pushed through two bags hoping to adjust but never did. The texture is good but that aftertaste is a dealbreaker.\",\n          \"The product itself is fine but the oil separation is the worst I have experienced in any natural nut butter. You need to stir for several minutes before every use and it still does not mix properly. Very frustrating.\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"HelpfulnessNumerator\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 18,\n        \"min\": 4,\n        \"max\": 78,\n        \"num_unique_values\": 74,\n        \"samples\": [\n          43,\n          63,\n          33\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"HelpfulnessDenominator\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 18,\n        \"min\": 12,\n        \"max\": 73,\n        \"num_unique_values\": 18,\n        \"samples\": [\n          36,\n          30,\n          65\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 5
        }
      ],
      "source": [
        "def load_fallback_sample():\n",
        "    \"\"\"Built-in Food & Grocery sample — loads if no CSV found.\"\"\"\n",
        "    print('⚠️  No CSV found — using built-in food reviews sample...')\n",
        "    records = [\n",
        "        {'ProductId':'F001','product_name':'Organic Green Tea','category':'Beverages','Score':5,\n",
        "         'Summary':'Best tea I have ever had',\n",
        "         'Text':'This organic green tea is absolutely wonderful. The flavour is delicate and fresh, nothing like the bitter supermarket brands. I have been drinking it every morning for 3 months and will never switch.','HelpfulnessNumerator':34,'HelpfulnessDenominator':36},\n",
        "        {'ProductId':'F001','product_name':'Organic Green Tea','category':'Beverages','Score':1,\n",
        "         'Summary':'Arrived stale and tasteless',\n",
        "         'Text':'Extremely disappointed. The tea had no aroma whatsoever and tasted like hot water. The bags were clearly old stock. I have tried to contact the seller with no response. Do not buy.','HelpfulnessNumerator':28,'HelpfulnessDenominator':30},\n",
        "        {'ProductId':'F002','product_name':'Dark Chocolate Bar 70%','category':'Confectionery','Score':5,\n",
        "         'Summary':'Incredible quality chocolate',\n",
        "         'Text':'This is hands down the best dark chocolate I have tasted outside of a specialist chocolatier. The bitterness is perfectly balanced with a hint of fruit. Melts beautifully. I order 6 bars at a time now.','HelpfulnessNumerator':52,'HelpfulnessDenominator':54},\n",
        "        {'ProductId':'F002','product_name':'Dark Chocolate Bar 70%','category':'Confectionery','Score':2,\n",
        "         'Summary':'Way too bitter, packaging damaged',\n",
        "         'Text':'The chocolate arrived with the wrapper partially open and one bar was broken into pieces. Even aside from that, 70% is too bitter for my taste. The description should be clearer about the intensity.','HelpfulnessNumerator':15,'HelpfulnessDenominator':18},\n",
        "        {'ProductId':'F003','product_name':'Extra Virgin Olive Oil','category':'Oils & Condiments','Score':5,\n",
        "         'Summary':'Authentic and fragrant',\n",
        "         'Text':'Finally found an olive oil that actually tastes like olive oil should. Peppery finish, grassy aroma, golden colour. I use it exclusively for salad dressings now. The quality is obvious from the first pour.','HelpfulnessNumerator':41,'HelpfulnessDenominator':43},\n",
        "        {'ProductId':'F003','product_name':'Extra Virgin Olive Oil','category':'Oils & Condiments','Score':3,\n",
        "         'Summary':'Decent but nothing special',\n",
        "         'Text':'It is a reasonable olive oil for everyday cooking but I would not call it extra virgin by taste. The aroma is mild and the flavour is neutral. Good value for the price but not premium quality.','HelpfulnessNumerator':9,'HelpfulnessDenominator':12},\n",
        "        {'ProductId':'F004','product_name':'Sourdough Crackers','category':'Snacks','Score':4,\n",
        "         'Summary':'Great with cheese, slightly salty',\n",
        "         'Text':'These crackers are excellent with cheese and charcuterie. The sourdough flavour is authentic and the texture is perfectly crispy. Only minor complaint is they are a bit salty on their own. Would buy again.','HelpfulnessNumerator':22,'HelpfulnessDenominator':23},\n",
        "        {'ProductId':'F004','product_name':'Sourdough Crackers','category':'Snacks','Score':1,\n",
        "         'Summary':'Half the box was broken crumbs',\n",
        "         'Text':'Terrible packaging. More than half the crackers arrived as crumbs because there is no internal protection in the box. The flavour was fine but completely unusable. I expect better for the price charged.','HelpfulnessNumerator':38,'HelpfulnessDenominator':40},\n",
        "        {'ProductId':'F005','product_name':'Manuka Honey 250g','category':'Spreads','Score':5,\n",
        "         'Summary':'Worth every penny for the health benefits',\n",
        "         'Text':'I have been taking a teaspoon of this every morning for my immune system and have noticed a real difference. The flavour is rich and distinctive. Authentic UMF rating, good traceability. Highly recommended.','HelpfulnessNumerator':63,'HelpfulnessDenominator':65},\n",
        "        {'ProductId':'F005','product_name':'Manuka Honey 250g','category':'Spreads','Score':2,\n",
        "         'Summary':'Not convinced it is genuine Manuka',\n",
        "         'Text':'For the price I expected more. The taste is sweeter than authentic Manuka should be and the UMF certification seems dubious. I have bought certified Manuka before and this does not compare. Very suspicious.','HelpfulnessNumerator':44,'HelpfulnessDenominator':48},\n",
        "        {'ProductId':'F006','product_name':'Instant Ramen Variety Pack','category':'Noodles','Score':4,\n",
        "         'Summary':'Great variety, solid flavour all round',\n",
        "         'Text':'Good selection of flavours in the pack. The spicy chicken is our household favourite. Quick to prepare and genuinely tasty for an instant product. The portion size is generous. Will subscribe for monthly delivery.','HelpfulnessNumerator':17,'HelpfulnessDenominator':18},\n",
        "        {'ProductId':'F006','product_name':'Instant Ramen Variety Pack','category':'Noodles','Score':2,\n",
        "         'Summary':'Far too much sodium, bland broth',\n",
        "         'Text':'The sodium content in these is absolutely alarming at over 1700mg per serving. The broth tastes more like salt water than anything else. I tried three flavours and they were all disappointing. Not worth it.','HelpfulnessNumerator':29,'HelpfulnessDenominator':31},\n",
        "        {'ProductId':'F007','product_name':'Cold Brew Coffee Concentrate','category':'Beverages','Score':5,\n",
        "         'Summary':'Smooth, rich and incredibly convenient',\n",
        "         'Text':'This cold brew concentrate has completely replaced my morning coffee shop habit. I just dilute 1:3 with milk and it is better than anything I can buy locally. Smooth, zero bitterness, great crema. Obsessed.','HelpfulnessNumerator':71,'HelpfulnessDenominator':73},\n",
        "        {'ProductId':'F007','product_name':'Cold Brew Coffee Concentrate','category':'Beverages','Score':1,\n",
        "         'Summary':'Leaked all over the box, appalling',\n",
        "         'Text':'The bottle seal failed in transit and the entire contents leaked destroying everything else in the delivery. There was no protective packaging around the bottle at all. This is basic logistics. Absolutely unacceptable.','HelpfulnessNumerator':55,'HelpfulnessDenominator':57},\n",
        "        {'ProductId':'F008','product_name':'Almond Butter Smooth','category':'Spreads','Score':4,\n",
        "         'Summary':'Clean ingredients, tastes great',\n",
        "         'Text':'Just almonds and a pinch of salt. No palm oil, no added sugar. The flavour is nutty and rich and it spreads well even straight from the fridge. Slightly expensive but the quality justifies it. Solid product.','HelpfulnessNumerator':26,'HelpfulnessDenominator':27},\n",
        "        {'ProductId':'F008','product_name':'Almond Butter Smooth','category':'Spreads','Score':3,\n",
        "         'Summary':'Oil separation is extreme and annoying',\n",
        "         'Text':'The product itself is fine but the oil separation is the worst I have experienced in any natural nut butter. You need to stir for several minutes before every use and it still does not mix properly. Very frustrating.','HelpfulnessNumerator':12,'HelpfulnessDenominator':14},\n",
        "        {'ProductId':'F009','product_name':'Protein Granola','category':'Breakfast','Score':5,\n",
        "         'Summary':'High protein and actually delicious',\n",
        "         'Text':'I have tried many protein granolas and they are usually chalky or overly sweet. This one is genuinely good. Crunchy clusters, real nuts, 20g protein per serving. Goes perfectly with Greek yoghurt. Reordering today.','HelpfulnessNumerator':33,'HelpfulnessDenominator':35},\n",
        "        {'ProductId':'F009','product_name':'Protein Granola','category':'Breakfast','Score':2,\n",
        "         'Summary':'Artificial sweetener aftertaste ruins it',\n",
        "         'Text':'The macros look great on paper but the stevia aftertaste is so strong it makes the whole bowl unpleasant. I pushed through two bags hoping to adjust but never did. The texture is good but that aftertaste is a dealbreaker.','HelpfulnessNumerator':21,'HelpfulnessDenominator':23},\n",
        "        {'ProductId':'F010','product_name':'Pasta Sauce Arrabbiata','category':'Sauces','Score':5,\n",
        "         'Summary':'Authentic Italian flavour',\n",
        "         'Text':'This sauce tastes like something from a Roman trattoria. The heat level is perfect, the tomatoes are clearly good quality and the whole thing is fragrant and fresh tasting. No unnecessary additives. Simply excellent.','HelpfulnessNumerator':48,'HelpfulnessDenominator':50},\n",
        "        {'ProductId':'F010','product_name':'Pasta Sauce Arrabbiata','category':'Sauces','Score':1,\n",
        "         'Summary':'Jar arrived broken, glass in the sauce',\n",
        "         'Text':'The jar cracked during delivery and when I opened it I could see glass fragments near the top. Clearly the courier handling was rough but there was also no protective padding in the box. Dangerous and unacceptable.','HelpfulnessNumerator':67,'HelpfulnessDenominator':68},\n",
        "    ]\n",
        "    extended = (records * 25)[:500]\n",
        "    df = pd.DataFrame(extended)\n",
        "    df['HelpfulnessNumerator'] = df['HelpfulnessNumerator'].apply(lambda x: max(0, x + np.random.randint(-5,8)))\n",
        "    print(f'✅ Built-in food sample loaded: {len(df)} rows')\n",
        "    return df\n",
        "\n",
        "# Load\n",
        "if os.path.exists(CSV_FILE_PATH):\n",
        "    df_raw = pd.read_csv(CSV_FILE_PATH, on_bad_lines='skip', engine='python', nrows=10000)\n",
        "    print(f'✅ Kaggle CSV loaded: {df_raw.shape}')\n",
        "    print(f'   Columns: {list(df_raw.columns)}')\n",
        "else:\n",
        "    df_raw = load_fallback_sample()\n",
        "\n",
        "df_raw.head(3)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_raw.to_csv(\"food_dataset.csv\", index=False)"
      ],
      "metadata": {
        "id": "4SCK_Wqew-pJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5PCVcj5JFGjS"
      },
      "source": [
        "---\n",
        "## 🧹 Step 4 — Data Cleaning & Standardisation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kFwPNZK5FGjS",
        "outputId": "42912beb-caea-45c9-cf6f-1ac722c301f0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ After cleaning: 500 rows × 8 columns\n",
            "\n",
            "Rating distribution:\n",
            "rating\n",
            "1    100\n",
            "2    100\n",
            "3     50\n",
            "4     75\n",
            "5    175\n"
          ]
        }
      ],
      "source": [
        "def clean_dataframe(df):\n",
        "    # Rename to standard schema (handles Amazon Fine Food columns)\n",
        "    col_map = {\n",
        "        'ProductId'             : 'product_id',\n",
        "        'Score'                 : 'rating',\n",
        "        'Summary'               : 'review_title',\n",
        "        'Text'                  : 'review_text',\n",
        "        'HelpfulnessNumerator'  : 'helpful_votes',\n",
        "        'HelpfulnessDenominator': 'total_votes',\n",
        "        'UserId'                : 'user_id',\n",
        "        'Time'                  : 'timestamp',\n",
        "        'ProfileName'           : 'reviewer_name',\n",
        "    }\n",
        "    df = df.rename(columns={k:v for k,v in col_map.items() if k in df.columns})\n",
        "\n",
        "    essential = ['product_id','rating','review_title','review_text']\n",
        "    optional  = ['product_name','category','helpful_votes','total_votes',\n",
        "                 'user_id','timestamp','reviewer_name']\n",
        "    keep = [c for c in essential + optional if c in df.columns]\n",
        "    df = df[keep].copy()\n",
        "\n",
        "    # Drop missing critical fields\n",
        "    df = df.dropna(subset=['review_text','rating'])\n",
        "    df['rating'] = pd.to_numeric(df['rating'], errors='coerce')\n",
        "    df = df[df['rating'].between(1,5)].copy()\n",
        "    df['rating'] = df['rating'].astype(int)\n",
        "\n",
        "    # Remove very short reviews\n",
        "    df = df[df['review_text'].astype(str).str.len() >= 40].copy()\n",
        "\n",
        "    # Clean text\n",
        "    df['review_text']  = df['review_text'].astype(str).str.replace('<br />', ' ').str.replace('<br/>', ' ').str.strip()\n",
        "    df['review_title'] = df['review_title'].fillna('No title').astype(str).str.strip()\n",
        "\n",
        "    # Helpful votes\n",
        "    if 'helpful_votes' in df.columns:\n",
        "        df['helpful_votes'] = pd.to_numeric(df['helpful_votes'], errors='coerce').fillna(0).astype(int)\n",
        "\n",
        "    return df.reset_index(drop=True)\n",
        "\n",
        "df_clean = clean_dataframe(df_raw)\n",
        "\n",
        "print(f'✅ After cleaning: {df_clean.shape[0]:,} rows × {df_clean.shape[1]} columns')\n",
        "print(f'\\nRating distribution:')\n",
        "print(df_clean['rating'].value_counts().sort_index().to_string())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ABdG_Rs8FGjS"
      },
      "source": [
        "---\n",
        "## 📊 Step 5 — Sample & Preview Raw Data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 451
        },
        "id": "gkrV0ZgZFGjS",
        "outputId": "06d5b756-84cb-45b4-a41c-74fdfa88c091"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "📋 Working sample: 425 reviews\n",
            "   Rating distribution: {1: 100, 2: 100, 3: 50, 4: 75, 5: 100}\n",
            "   Unique products    : 10\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/tmp/ipykernel_2503/3507263612.py:35: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  plt.tight_layout()\n",
            "/tmp/ipykernel_2503/3507263612.py:36: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  plt.savefig('01_raw_data_overview.png', dpi=150, bbox_inches='tight')\n",
            "/usr/local/lib/python3.12/dist-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  fig.canvas.print_figure(bytes_io, **kw)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1300x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "📊 Saved: 01_raw_data_overview.png\n"
          ]
        }
      ],
      "source": [
        "# Stratified sample — equal reviews per star rating for balanced analysis\n",
        "per_rating = N_SAMPLE // 5\n",
        "df_sample = pd.concat([\n",
        "    df_clean[df_clean['rating']==s].sample(\n",
        "        n=min(per_rating, (df_clean['rating']==s).sum()),\n",
        "        random_state=RANDOM_SEED\n",
        "    )\n",
        "    for s in [1,2,3,4,5]\n",
        "]).sample(frac=1, random_state=RANDOM_SEED).reset_index(drop=True)\n",
        "\n",
        "print(f'📋 Working sample: {len(df_sample)} reviews')\n",
        "print(f'   Rating distribution: {df_sample[\"rating\"].value_counts().sort_index().to_dict()}')\n",
        "print(f'   Unique products    : {df_sample[\"product_id\"].nunique()}')\n",
        "\n",
        "fig, axes = plt.subplots(1, 2, figsize=(13, 5))\n",
        "fig.suptitle('Raw Data Overview — Amazon Fine Food Reviews', fontsize=14, fontweight='bold')\n",
        "\n",
        "df_sample['rating'].value_counts().sort_index().plot(\n",
        "    kind='bar', ax=axes[0],\n",
        "    color=['#e74c3c','#e67e22','#f1c40f','#2ecc71','#27ae60'], edgecolor='white'\n",
        ")\n",
        "axes[0].set_title('Rating Distribution (Sample)', fontweight='bold')\n",
        "axes[0].set_xticklabels(['1⭐','2⭐','3⭐','4⭐','5⭐'], rotation=0)\n",
        "axes[0].set_ylabel('Number of Reviews')\n",
        "for p in axes[0].patches:\n",
        "    axes[0].annotate(str(int(p.get_height())),\n",
        "                     (p.get_x()+p.get_width()/2, p.get_height()),\n",
        "                     ha='center', va='bottom', fontsize=10)\n",
        "\n",
        "rule_sent = df_sample['rating'].map({1:'Negative',2:'Negative',3:'Neutral',4:'Positive',5:'Positive'}).value_counts()\n",
        "axes[1].pie(rule_sent, labels=rule_sent.index, autopct='%1.1f%%', startangle=90,\n",
        "            colors=[{'Positive':'#27ae60','Neutral':'#f39c12','Negative':'#e74c3c'}.get(s,'grey') for s in rule_sent.index])\n",
        "axes[1].set_title('Baseline Sentiment (from Star Rating)', fontweight='bold')\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.savefig('01_raw_data_overview.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "print('📊 Saved: 01_raw_data_overview.png')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_36yXm5AFGjS"
      },
      "source": [
        "---\n",
        "## 🤖 Step 6 — TextBlob Sentiment Enrichment\n",
        "No API needed — runs locally and instantly."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "do-M5kLNFGjS",
        "outputId": "463075b2-d481-4bc7-b087-a22f88033dfa"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "🔍 Computing TextBlob sentiment...\n",
            "✅ Done — 425 reviews labelled\n",
            "\n",
            "Sentiment distribution:\n",
            "sentiment_label\n",
            "Positive    207\n",
            "Neutral     168\n",
            "Negative     50\n"
          ]
        }
      ],
      "source": [
        "def get_sentiment(text):\n",
        "    score = round(TextBlob(str(text)).sentiment.polarity, 3)\n",
        "    if score > 0.1:   label = 'Positive'\n",
        "    elif score < -0.1: label = 'Negative'\n",
        "    else:              label = 'Neutral'\n",
        "    return label, score\n",
        "\n",
        "print('🔍 Computing TextBlob sentiment...')\n",
        "results = df_sample['review_text'].apply(get_sentiment)\n",
        "df_sample['sentiment_label'] = results.apply(lambda x: x[0])\n",
        "df_sample['sentiment_score'] = results.apply(lambda x: x[1])\n",
        "\n",
        "print(f'✅ Done — {len(df_sample)} reviews labelled')\n",
        "print(f'\\nSentiment distribution:')\n",
        "print(df_sample['sentiment_label'].value_counts().to_string())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lcud7PwIFGjS"
      },
      "source": [
        "---\n",
        "## 🧬 Step 7 — Synthetic Data Generation\n",
        "Pre-written food review pools (same approach as class notebook).\n",
        "Generated with LLM assistance and hardcoded — **no API calls**."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "m9KLWGrgFGjS",
        "outputId": "ea5f9654-ff90-47aa-d73f-17401d289969"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Food synthetic review pools ready\n",
            "   Positive: 20 reviews\n",
            "   Neutral: 20 reviews\n",
            "   Negative: 20 reviews\n"
          ]
        }
      ],
      "source": [
        "# ── FOOD-SPECIFIC SYNTHETIC REVIEW POOLS ─────────────────────────\n",
        "# Created by prompting an LLM:\n",
        "# \"Write 20 realistic Amazon food product reviews that are [positive/neutral/negative].\n",
        "#  Vary the products: snacks, beverages, condiments, health foods.\"\n",
        "\n",
        "synthetic_reviews_by_sentiment = {\n",
        "\n",
        "    'Positive': [\n",
        "        \"Absolutely the best version of this product I have ever tried. The flavour is authentic and the quality is clearly a step above the supermarket equivalent.\",\n",
        "        \"I ordered this on a whim and it has become a household staple. We go through a pack every week. The taste is consistent and the price is fair for the quality.\",\n",
        "        \"Exceptional product. Fresh, well packaged and exactly as described. The flavour profile is complex and satisfying. Will definitely be reordering regularly.\",\n",
        "        \"This exceeded my expectations completely. I was sceptical but the reviews were right. Outstanding flavour and the ingredients list is clean and simple.\",\n",
        "        \"Best purchase I have made on this platform. The product arrived quickly, was well protected and tasted incredible straight out of the packaging.\",\n",
        "        \"I have tried many similar products and this is head and shoulders above the competition. Rich, full flavoured and zero artificial aftertaste.\",\n",
        "        \"Five stars without hesitation. This is the kind of quality you expect from a specialist supplier. The freshness is obvious and the taste is superb.\",\n",
        "        \"My whole family loves this. We have been buying it for six months and the quality has been absolutely consistent every single time. Highly recommend.\",\n",
        "        \"Genuinely delicious. I bought this as a gift and the recipient immediately asked where to get more. That says everything about the quality.\",\n",
        "        \"Perfect product. The packaging keeps everything fresh, the flavour is exactly right and the portion size is generous. Great value for money.\",\n",
        "        \"I am so glad I found this. It has transformed the way I cook and the flavour it adds is remarkable. Worth every penny and I will never go back.\",\n",
        "        \"Outstanding quality from start to finish. The aroma when you open the packaging is wonderful and the taste lives up to it completely.\",\n",
        "        \"This is now my go-to product in this category. Consistent quality, brilliant taste and it keeps well. Cannot recommend it highly enough.\",\n",
        "        \"Honestly surprised by how good this is. The description does not do it justice. Rich, satisfying and made with ingredients you can actually read.\",\n",
        "        \"Subscribed after the first order. The value is excellent for the quality and my morning routine has genuinely improved since I started using it.\",\n",
        "        \"Fantastic product that delivers on every single claim on the label. Fresh, flavourful and made the right way. A cut above everything else I have tried.\",\n",
        "        \"The whole household is obsessed. We have already reordered three times. Great flavour, clean label and arrives perfectly packaged every time.\",\n",
        "        \"Exactly what I was looking for. High quality ingredients, honest flavour and no unnecessary additives. This is how food should be made.\",\n",
        "        \"Brilliant find. I bought it based on reviews and the reviews were absolutely right. Delicious, fresh and great for the price. Will reorder always.\",\n",
        "        \"A product that genuinely lives up to its reputation. The quality is premium, the taste is exceptional and the packaging is thoughtful and secure.\",\n",
        "    ],\n",
        "\n",
        "    'Neutral': [\n",
        "        \"Decent product that does what it says. Nothing groundbreaking but a reliable choice. I would buy it again if I could not find anything better locally.\",\n",
        "        \"It is fine. The flavour is acceptable and the quality is what you would expect at this price point. Not remarkable but not disappointing either.\",\n",
        "        \"Average experience overall. The product is edible and the packaging is adequate. I expected more based on the description but it is not bad.\",\n",
        "        \"Mixed feelings. Some of the product was great and some was underwhelming. Inconsistent quality across the batch which was a bit frustrating.\",\n",
        "        \"It does the job. Not the most exciting product in the category but reliable and practical. A reasonable everyday option at a fair price.\",\n",
        "        \"Three stars because it is neither impressive nor terrible. Solid middle ground product. Functional and consistent if not particularly memorable.\",\n",
        "        \"I have bought better and I have bought worse. This sits comfortably in the middle. Worth trying but manage your expectations accordingly.\",\n",
        "        \"The flavour is pleasant enough without being outstanding. Would work well as part of a recipe but perhaps not as a standalone snack.\",\n",
        "        \"Not sure if I would reorder. The quality is acceptable but there is nothing that distinguishes this from a cheaper supermarket alternative.\",\n",
        "        \"Reasonable product for the price. Arrived in good condition and tasted fine. Just nothing special that would make me prioritise it over other options.\",\n",
        "        \"Okay but not memorable. The ingredients are fine, the texture is acceptable, the flavour is inoffensive. A three-star product in every sense.\",\n",
        "        \"Does what it says on the label, no more and no less. I appreciate the clean ingredients but the overall experience was a bit flat.\",\n",
        "        \"Perfectly serviceable. I used it up without complaint but also without enthusiasm. There are probably better options in this category if you look around.\",\n",
        "        \"Average quality, average taste, average experience. Not a bad product but certainly not a great one. Probably will not repurchase at this price.\",\n",
        "        \"It is fine for what it is. The convenience factor is decent and the flavour is acceptable. Just do not expect anything extraordinary.\",\n",
        "        \"A middle-of-the-road purchase. Not the worst thing I have bought and not the best. Gets the job done without exciting or disappointing.\",\n",
        "        \"The product is what it is. Good enough for regular use but I would not recommend it to anyone looking for something exceptional.\",\n",
        "        \"Three stars feels right. Competent product, no major issues, but also no reason to choose it over competitors unless the price is right.\",\n",
        "        \"Perfectly adequate. There is nothing wrong with it but there is also nothing that makes it stand out. A background product rather than a star.\",\n",
        "        \"It gets the job done. Taste is neutral, texture is acceptable, packaging is functional. Nothing to rave about but nothing to complain about either.\",\n",
        "    ],\n",
        "\n",
        "    'Negative': [\n",
        "        \"Extremely disappointed. The product looked nothing like the photos and the taste was completely off. This does not match the description at all.\",\n",
        "        \"Arrived in terrible condition. The packaging was crushed and the product inside was damaged beyond use. No protective padding whatsoever.\",\n",
        "        \"Worst food purchase I have made online. The smell when I opened it was immediately wrong. Clearly old stock being sold as fresh. Do not buy.\",\n",
        "        \"The product is fine but the portion size is a complete joke for the price. I cannot believe the amount you actually get. Total rip off.\",\n",
        "        \"Flavour is artificial and overwhelming. Nothing like the natural product it claims to be. The ingredient list does not match the front of the pack.\",\n",
        "        \"I received the wrong item and when I contacted support the response was dismissive and unhelpful. Still waiting for a resolution two weeks later.\",\n",
        "        \"Mouldy on arrival. Clearly stored badly before dispatch. The best before date had not passed but the product was inedible. Shocking quality control.\",\n",
        "        \"The sodium level in this product is dangerous. Over 2000mg per serving is not something that should be on shelves without a clear health warning.\",\n",
        "        \"Absolutely terrible. I followed the preparation instructions exactly and the result was inedible. I cannot understand how this product has positive reviews.\",\n",
        "        \"Leaking packaging meant the product was ruined and contaminated other items in the delivery. Completely unacceptable and a waste of money.\",\n",
        "        \"Do not believe the health claims on this packaging. They are wildly misleading and the actual nutritional content tells a very different story.\",\n",
        "        \"The quality has clearly dropped since previous orders. Same product but completely different taste. Either the recipe has changed or the ingredients.\",\n",
        "        \"I have been sick after eating this product. I cannot prove it was the cause but the timing and symptoms were very suspicious. Removing from my diet.\",\n",
        "        \"Ordered twice hoping the first bad experience was a fluke. Both times the same problem. Clearly a systemic issue with this product. Giving up.\",\n",
        "        \"The product contains allergens not listed on the website description. This is a serious safety issue that needs to be addressed immediately.\",\n",
        "        \"Stale, tasteless and overpriced. I have bought the supermarket version for a third of the price and it is genuinely better. Complete waste of money.\",\n",
        "        \"Customer service was appalling when I reported the issue. Dismissive, slow and ultimately unhelpful. Will not be using this seller again.\",\n",
        "        \"The texture is completely wrong for this type of product. It was gummy and unpleasant where it should have been light and crisp. Inedible.\",\n",
        "        \"One star because zero is not an option. Every aspect of this purchase was a disappointment from the misleading photos to the awful flavour.\",\n",
        "        \"Returned immediately. The product was clearly not as advertised and the quality was so poor that I questioned whether it was genuine at all.\",\n",
        "    ]\n",
        "}\n",
        "\n",
        "print('✅ Food synthetic review pools ready')\n",
        "for s, r in synthetic_reviews_by_sentiment.items():\n",
        "    print(f'   {s}: {len(r)} reviews')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "PpvNBJpZFGjT",
        "outputId": "3f8f6b0e-876c-4483-9a10-2271c28b91af"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "✅ Generated 60 synthetic food reviews across 6 products\n",
            "\n",
            "Synthetic sentiment distribution:\n",
            "sentiment_label\n",
            "Positive    30\n",
            "Negative    20\n",
            "Neutral     10\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "            product_name   category  rating sentiment_label  \\\n",
              "0  Organic Matcha Powder  Beverages       5        Positive   \n",
              "1  Organic Matcha Powder  Beverages       5        Positive   \n",
              "2  Organic Matcha Powder  Beverages       5        Positive   \n",
              "3  Organic Matcha Powder  Beverages       5        Positive   \n",
              "4  Organic Matcha Powder  Beverages       5        Positive   \n",
              "\n",
              "                                                                                                               review_text  \n",
              "0  This exceeded my expectations completely. I was sceptical but the reviews were right. Outstanding flavour and the in...  \n",
              "1  Absolutely the best version of this product I have ever tried. The flavour is authentic and the quality is clearly a...  \n",
              "2  Genuinely delicious. I bought this as a gift and the recipient immediately asked where to get more. That says everyt...  \n",
              "3  My whole family loves this. We have been buying it for six months and the quality has been absolutely consistent eve...  \n",
              "4  The whole household is obsessed. We have already reordered three times. Great flavour, clean label and arrives perfe...  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-636ca103-aba9-47a9-b5cd-9fa868e29d72\" 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>product_name</th>\n",
              "      <th>category</th>\n",
              "      <th>rating</th>\n",
              "      <th>sentiment_label</th>\n",
              "      <th>review_text</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>Organic Matcha Powder</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Positive</td>\n",
              "      <td>This exceeded my expectations completely. I was sceptical but the reviews were right. Outstanding flavour and the in...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Organic Matcha Powder</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Positive</td>\n",
              "      <td>Absolutely the best version of this product I have ever tried. The flavour is authentic and the quality is clearly a...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Organic Matcha Powder</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Positive</td>\n",
              "      <td>Genuinely delicious. I bought this as a gift and the recipient immediately asked where to get more. That says everyt...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Organic Matcha Powder</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Positive</td>\n",
              "      <td>My whole family loves this. We have been buying it for six months and the quality has been absolutely consistent eve...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Organic Matcha Powder</td>\n",
              "      <td>Beverages</td>\n",
              "      <td>5</td>\n",
              "      <td>Positive</td>\n",
              "      <td>The whole household is obsessed. We have already reordered three times. Great flavour, clean label and arrives perfe...</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-636ca103-aba9-47a9-b5cd-9fa868e29d72')\"\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-636ca103-aba9-47a9-b5cd-9fa868e29d72 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-636ca103-aba9-47a9-b5cd-9fa868e29d72');\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>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "summary": "{\n  \"name\": \"df_synthetic[['product_name','category','rating','sentiment_label','review_text']]\",\n  \"rows\": 5,\n  \"fields\": [\n    {\n      \"column\": \"product_name\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Organic Matcha Powder\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"category\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Beverages\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"rating\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 5,\n        \"max\": 5,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          5\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"sentiment_label\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 1,\n        \"samples\": [\n          \"Positive\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"review_text\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 5,\n        \"samples\": [\n          \"Absolutely the best version of this product I have ever tried. The flavour is authentic and the quality is clearly a step above the supermarket equivalent.\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 10
        }
      ],
      "source": [
        "# ── SYNTHETIC FOOD PRODUCTS ───────────────────────────────────────\n",
        "# Underrepresented product/scenario combinations not in the Kaggle data\n",
        "synthetic_products = [\n",
        "    {'product_id':'SYN001','product_name':'Organic Matcha Powder',     'category':'Beverages',        'rating':5,'sentiment_label':'Positive', 'n_reviews':10},\n",
        "    {'product_id':'SYN002','product_name':'Gluten-Free Pasta',          'category':'Pasta & Grains',   'rating':2,'sentiment_label':'Negative', 'n_reviews':10},\n",
        "    {'product_id':'SYN003','product_name':'Truffle Sea Salt Flakes',    'category':'Oils & Condiments','rating':4,'sentiment_label':'Positive', 'n_reviews':10},\n",
        "    {'product_id':'SYN004','product_name':'Vegan Protein Bar',          'category':'Health Foods',     'rating':3,'sentiment_label':'Neutral',  'n_reviews':10},\n",
        "    {'product_id':'SYN005','product_name':'Raw Cashew Butter',          'category':'Spreads',          'rating':1,'sentiment_label':'Negative', 'n_reviews':10},\n",
        "    {'product_id':'SYN006','product_name':'Smoked Paprika Almonds',     'category':'Snacks',           'rating':5,'sentiment_label':'Positive', 'n_reviews':10},\n",
        "]\n",
        "\n",
        "# Generate synthetic reviews — same logic as class notebook (random.sample)\n",
        "synthetic_rows = []\n",
        "for product in synthetic_products:\n",
        "    pool    = synthetic_reviews_by_sentiment[product['sentiment_label']]\n",
        "    sampled = random.sample(pool, product['n_reviews'])\n",
        "    for review_text in sampled:\n",
        "        _, score = get_sentiment(review_text)\n",
        "        synthetic_rows.append({\n",
        "            'product_id'      : product['product_id'],\n",
        "            'product_name'    : product['product_name'],\n",
        "            'category'        : product['category'],\n",
        "            'rating'          : product['rating'],\n",
        "            'review_title'    : product['sentiment_label'] + ' experience',\n",
        "            'review_text'     : review_text,\n",
        "            'sentiment_label' : product['sentiment_label'],\n",
        "            'sentiment_score' : score,\n",
        "            'helpful_votes'   : random.randint(0, 60),\n",
        "            'total_votes'     : random.randint(0, 70),\n",
        "            'is_synthetic'    : True\n",
        "        })\n",
        "\n",
        "df_synthetic = pd.DataFrame(synthetic_rows)\n",
        "print(f'✅ Generated {len(df_synthetic)} synthetic food reviews across {len(synthetic_products)} products')\n",
        "print(f'\\nSynthetic sentiment distribution:')\n",
        "print(df_synthetic['sentiment_label'].value_counts().to_string())\n",
        "df_synthetic[['product_name','category','rating','sentiment_label','review_text']].head(5)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HyfLzpDOFGjT"
      },
      "source": [
        "---\n",
        "## 💾 Step 8 — Combine Real + Synthetic & Save"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vDyyg1LXFGjT",
        "outputId": "cac9b6f5-cee5-423a-f21f-52043c85d702"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "📦 Final dataset:\n",
            "   Real reviews (Kaggle) : 425\n",
            "   Synthetic reviews     : 60\n",
            "   Total rows            : 485\n",
            "   Unique real products  : 10\n",
            "   Synthetic products    : 6\n",
            "\n",
            "💾 Saved to: food_reviews_enriched.csv\n",
            "   ➡️  Load this in Notebook 2!\n"
          ]
        }
      ],
      "source": [
        "df_sample['is_synthetic'] = False\n",
        "df_final = pd.concat([df_sample, df_synthetic], ignore_index=True, sort=False)\n",
        "\n",
        "# Final type cleanup\n",
        "df_final['rating']          = pd.to_numeric(df_final['rating'], errors='coerce').fillna(3).astype(int)\n",
        "df_final['sentiment_score'] = pd.to_numeric(df_final['sentiment_score'], errors='coerce').fillna(0.0).round(3)\n",
        "df_final['helpful_votes']   = pd.to_numeric(df_final['helpful_votes'], errors='coerce').fillna(0).astype(int)\n",
        "\n",
        "OUTPUT = 'food_reviews_enriched.csv'\n",
        "df_final.to_csv(OUTPUT, index=False)\n",
        "\n",
        "print('📦 Final dataset:')\n",
        "print(f'   Real reviews (Kaggle) : {len(df_sample)}')\n",
        "print(f'   Synthetic reviews     : {len(df_synthetic)}')\n",
        "print(f'   Total rows            : {len(df_final)}')\n",
        "print(f'   Unique real products  : {df_sample[\"product_id\"].nunique()}')\n",
        "print(f'   Synthetic products    : {df_synthetic[\"product_id\"].nunique()}')\n",
        "print(f'\\n💾 Saved to: {OUTPUT}')\n",
        "print(f'   ➡️  Load this in Notebook 2!')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 561
        },
        "id": "9rJ5UnRwFGjT",
        "outputId": "cda58d33-f15a-41f0-bd20-62935587e6ba"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/tmp/ipykernel_2503/3320424086.py:39: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  plt.tight_layout()\n",
            "/tmp/ipykernel_2503/3320424086.py:40: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  plt.savefig('02_enriched_summary.png', dpi=150, bbox_inches='tight')\n",
            "/usr/local/lib/python3.12/dist-packages/IPython/core/pylabtools.py:151: UserWarning: Glyph 11088 (\\N{WHITE MEDIUM STAR}) missing from font(s) Liberation Sans.\n",
            "  fig.canvas.print_figure(bytes_io, **kw)\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1400x1000 with 4 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "📊 Saved: 02_enriched_summary.png\n"
          ]
        }
      ],
      "source": [
        "# Final summary charts\n",
        "fig, axes = plt.subplots(2, 2, figsize=(14, 10))\n",
        "fig.suptitle('Final Enriched Dataset — Amazon Food Reviews', fontsize=15, fontweight='bold')\n",
        "\n",
        "# 1. Sentiment distribution\n",
        "sc = df_final['sentiment_label'].value_counts()\n",
        "clr = [{'Positive':'#27ae60','Neutral':'#f39c12','Negative':'#e74c3c'}.get(s,'grey') for s in sc.index]\n",
        "sc.plot(kind='bar', ax=axes[0,0], color=clr, edgecolor='white', width=0.6)\n",
        "axes[0,0].set_title('Sentiment Distribution', fontweight='bold')\n",
        "axes[0,0].set_xticklabels(sc.index, rotation=0)\n",
        "for p in axes[0,0].patches:\n",
        "    axes[0,0].annotate(str(int(p.get_height())),\n",
        "                       (p.get_x()+p.get_width()/2, p.get_height()),\n",
        "                       ha='center', va='bottom', fontsize=10)\n",
        "\n",
        "# 2. Real vs Synthetic\n",
        "tc = df_final['is_synthetic'].map({False:'Real',True:'Synthetic'}).value_counts()\n",
        "axes[0,1].pie(tc, labels=tc.index, autopct='%1.1f%%', startangle=90,\n",
        "              colors=['#3498db','#e67e22'])\n",
        "axes[0,1].set_title('Real vs Synthetic Reviews', fontweight='bold')\n",
        "\n",
        "# 3. Rating distribution\n",
        "df_final['rating'].value_counts().sort_index().plot(\n",
        "    kind='bar', ax=axes[1,0],\n",
        "    color=['#e74c3c','#e67e22','#f1c40f','#2ecc71','#27ae60'], edgecolor='white')\n",
        "axes[1,0].set_title('Rating Distribution (Combined)', fontweight='bold')\n",
        "axes[1,0].set_xticklabels(['1⭐','2⭐','3⭐','4⭐','5⭐'], rotation=0)\n",
        "\n",
        "# 4. Sentiment score histogram\n",
        "df_final['sentiment_score'].plot(\n",
        "    kind='hist', bins=25, ax=axes[1,1], color='#9b59b6', edgecolor='white', alpha=0.85)\n",
        "axes[1,1].axvline(x=0, color='black', linestyle='--', alpha=0.6, label='Neutral (0)')\n",
        "axes[1,1].axvline(df_final['sentiment_score'].mean(), color='red', linestyle='--', alpha=0.7,\n",
        "                  label=f'Mean ({df_final[\"sentiment_score\"].mean():.2f})')\n",
        "axes[1,1].set_title('Sentiment Score Distribution (TextBlob)', fontweight='bold')\n",
        "axes[1,1].set_xlabel('-1.0 = Very Negative  →  +1.0 = Very Positive')\n",
        "axes[1,1].legend()\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.savefig('02_enriched_summary.png', dpi=150, bbox_inches='tight')\n",
        "plt.show()\n",
        "print('📊 Saved: 02_enriched_summary.png')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fl3ecHwoFGjT"
      },
      "source": [
        "---\n",
        "## ✅ Notebook 1 Complete!\n",
        "\n",
        "| Step | Output |\n",
        "|---|---|\n",
        "| Real data | Amazon Fine Food Reviews (Kaggle) — properly filtered |\n",
        "| Synthetic data | 60 food reviews across 6 new products |\n",
        "| Sentiment | TextBlob labels + scores on every review |\n",
        "| **Output file** | `food_reviews_enriched.csv` |\n",
        "| **Cost** | **$0.00** |\n",
        "\n",
        "### ➡️ Next: Open `notebook_2_analysis.ipynb` and load `food_reviews_enriched.csv`"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.10.0"
    },
    "colab": {
      "provenance": [],
      "toc_visible": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}