{
  "fetched_at": "2026-10-20T00:59:07Z",
  "source_url": "https://www.kaggle.com/competitions/2022tae/data",
  "pages": [
    {
      "name": "data-description",
      "content": "##Files\nAll teams are provided with the following files - training dataset (**train.csv**) and test dataset (**test.csv**).\n\n##Training dataset\n- ```tweet``` - (character) a string of letters and symbols representing a tweet\n- ```sentiment``` - (integer) an integer representing a sentiment corresponding to a tweet: 1 indicates *negative* sentiment, 2 indicates *neutral*, 3 indicates *positive*\n\n##Test dataset\n- ```id``` - (integer) id corresponding to a tweet in the test dataset\n- ```tweet``` - (character) a string of letters and symbols representing a tweet which sentiment needs to be predicted\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "## Your Task\nEach team is to develop an algorithm that determines what sort of sentiment a given set of tweet references - does the tweet have a *negative*, *neutral* or *positive* sentiment?\n\n## Datasets\nThe following datasets are provided:\n\n- **train.csv**: 20,610 tweets with the corresponding classification (sentiment) - integers \"1\", \"2\" and \"3\" indicate *negative*, *neutral* and *positive* sentiment, respectively.\n- **test.csv**: 6,871 tweets with no labels - to be used for quantifying the performance of the algorithms.\n\n## Rules\nPlease refer to the tab on 'Rules' and class slides provided on eDimension for more details.\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "##Performance Metric\nThe performance of the algorithms will be evaluated based on *classification accuracy* (the ratio between the number of correctly classified samples to the total number of samples). Randomly assigning labels to the tweets in the test dataset gives an accuracy of approximately 30%.\n\n##Leaderboards\nKaggle will calculate the accuracy on two subsets of the test dataset - *public* and *private*. The *public* leaderboard will be available during the competition while the *private* leaderboard will only be available at the end of the competition.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "##Code and Methods\n- Only R is allowed, but you may use any R package\n- Methods not covered in class are allowed (e.g. neural networks)\n\n##Teams\n- One Kaggle account per team (your team cannot submit from multiple Kaggle accounts)\n- Team mergers are not allowed\n\n##Submissions\n- Each team may submit a maximum of 2 entries per day\n- Each team may select only 1 final submission for judging",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Submission (Kaggle)",
      "content": "##Submission Format\nSubmit a .csv with two columns named ```id``` and ```sentiment```. Sticking with the exact capitalisation is advised as these headers must be present for evaluation.\n\nEach entry of ```id``` corresponds to a single tweet as given in the test dataset **test.csv** with ```id``` sequentially increasing from 1 to 6871.\n\nThe ```sentiment``` is the sentiment label as predicted by you, an integer of 1, 2 or 3 (check that they are integers).\n\nThe first 5 lines of a .csv submission should have a format like the following:\n| id   | sentiment |\n| --- | ---             |\n|     1 |               3 |\n|     2 |               2 |\n|     3 |               2 |\n|     4|                 1|\n|     5 |               3 |\n\nEach team can submit up to 2 entries a day. The final day of submission is Aug 11, 23:59.##",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Submission (eDimension)",
      "content": "Each team is to submit on eDimension a report along with your code and a README file (in a zipped folder).\n\nThe report is a short document (maximum 4 pages, font size 12) containing:\n\n- a high-level description of the approach developed\n- a short description of the results\n- a brief discussion on interpretability and limits of the approach\n\nThe README file should contain a list of all dependencies and files need to run your code. Do note that your code must produce the same results as what you have uploaded on Kaggle.\n\nThe reports and code are due on Aug 14 23:59 (SGT).",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Schedule",
      "content": "##Competition Schedule\n\n| Date  | Event |\n| --- | --- |\n| August 3, 2022 | Announcement of the Data Competition  |\n| August 3, 2022 15:00| Publication of problem details and competition rules |\n| August 11, 2022 23:59 | Last opportunity for submitting the results on Kaggle |\n| August 14, 2022 23:59 | Submission of reports, code, and peer evaluation form on eDimension|",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}