{
  "fetched_at": "2026-10-19T20:41:41Z",
  "source_url": "https://www.kaggle.com/competitions/TweetSentimentBR/data",
  "pages": [
    {
      "name": "data-description",
      "content": "We will have two sets of data in csv format, the data for training the model, with 50000 lines, and the data for classification with 5000 lines. The main variable you should sort on will be `sentiment`.\n\n##Files\n\n* **Train.csv** – the training set\n* **Test.csv** – the test suite\n* **Sample_Submission.csv** – a sample submission file in the correct format\n\n## Columns\n\n- `id`: Integer\n- `tweet_text`: String\n- `tweet_date`: Date and time\n- `sentiment`: Categorical (`1` for Positive, `0` for Negative)\n- `query_used`: Categorical ( `:(`, `:)` )",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "For this competition, I will provide the dataset for analyzing the sentiments of Brazilians on Twitter, here we have a collection of tweets that were analyzed to determine **Positive** `:)` or **Negative** `:(` sentiment. This dataset is a valuable tool for participants, allowing them to develop and test sentiment analysis models in a realistic environment.\n\n### The dataset is composed of five main columns:\n\n1. id: A unique identifier for each tweet\n2. tweet_text: The text of the tweet. This field contains the content of the tweet\n3. tweet_date: The date and time the tweet was posted. This field can be useful for analyzing temporal trends or understanding the context in which the tweet was written.\n4. sentiment: The sentiment label of the tweet, which indicates whether the sentiment expressed is Positive `1`, Negative `0`. This field is the target variable for sentiment analysis. **Tweets were modified to remove the characters `:)` and `:(` so as not to be obvious for model classification**\n5. query_used: Character used for tracking tweet `:)` and `:(`\n\n#### Data Types:\n- `id`: Integer\n- `tweet_text`: String\n- `tweet_date`: Date and time\n- `sentiment`: Categorical (`1` for Positive, `0` for Negative)\n- `query_used`: Categorical ( `:(`, `:)` )",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "Participants must use the dataset to develop machine learning models that can predict the sentiment expressed in tweets. They must train their models using the provided training set and submit their predictions to the test set for evaluation. Good luck\n\n---\n\nParticipants will be evaluated based on the accuracy of their sentiment predictions on tweets from the provided test set\n\n\n## Submission File\nThe file must contain a header and have the following format:\n| id | sentiment |\n| --- | --- |\n| 2129535486121049552 | 1 |\n| 2130527465221556147 | 0 |",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "> ### Let's Go\n- Don't cheat!\n- Apply yourself!\n- Have fun :)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "abstract",
      "content": "Welcome to the Tweet Sentiment Analysis Challenge! This challenge aims to assess participants' ability to analyze and predict Brazilians' sentiments in tweets using machine learning and natural language processing (NLP) techniques. Participants will be challenged to develop models that can correctly identify and classify sentiment expressed in a wide range of tweets.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}