{
  "fetched_at": "2026-10-20T06:34:40Z",
  "source_url": "https://www.kaggle.com/competitions/fit5149-tp6-2021-a2/data",
  "pages": [
    {
      "name": "data-description",
      "content": "# Data\n\nWe provide the following data sets: \n- trainset_students.csv contains training IDs, labels, and labels confidence. It contains more than 9,000 images and acts as the training data.\n- ID: the unique image identifier. The actual corresponding image can be found here [link] \nLabel: the image sentiment polarity score. The labels are highly positive, positive, neutral, negative, and highly negative.\n- Label_confidence: The confidence of the image Label between 0 to 1 where 1 indicates 100% confidence over the Label. Please note that each image is manually labeled by several human annotators and this column indicates their agreement with the image sentiment polarity.\n- testset_students.csv: only testing ids are available. It contains 3146 images. The actual corresponding image can be found here [link] ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "In this assessment, you are given a data set that contains over twelve thousand sentiment-scored images. Contributors were shown a variety of pictures (everything from portraits of celebrities to landscapes to stock photography) and asked to score the images on typical positive/negative sentiment. The data set contains the actual images (stored and shared with you in Google Drive), sentiment scores of highly positive, positive, neutral, negative, and highly negative, and contributor agreement. The task is to develop classification models which can make predictions and return the corresponding classification of the source image. There are many machine learning methods that can be used in the classification task and you can use any one of them. Figure 1 shows a typical framework used in the supervised classification.\n\n# Data\nWe provide the following data sets: \n- trainset_students.csv contains training IDs, labels, and labels confidence. It contains more than 9,000 images and acts as the training data.\n- ID: the unique image identifier. The actual corresponding image can be found here [link] \nLabel: the image sentiment polarity score. The labels are highly positive, positive, neutral, negative, and highly negative.\n- Label_confidence: The confidence of the image Label between 0 to 1 where 1 indicates 100% confidence over the Label. Please note that each image is manually labeled by several human annotators and this column indicates their agreement with the image sentiment polarity.\n- testset_students.csv: only testing ids are available. It contains 3146 images. The actual corresponding image can be found here [link] ",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "The metric to be used in this challenge is Accuracy. A brief description of this metric can be found [here](https://en.wikipedia.org/wiki/Accuracy_and_precision).\n\n# Submission Format\nFor every image in the dataset, submission files should contain two columns: ID and Label.\n\nThe file should contain a header and have the following format:\n\n```\nID, Label\n0, Positive\n1, Negative\netc.\n```",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "This is a page where you can include rules that participants must accept before joining. You may wish to include rules like:\n\n - Don't cheat!\n - Apply yourself!\n - Have fun!",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}