{
  "id": 405460,
  "title": "Do people find it useful to use the raw data?",
  "url": "/competitions/predict-student-performance-from-game-play/discussion/405460",
  "author_name": "",
  "post_date": "2023-04-27T17:19:45.864446900Z",
  "votes": 3,
  "comment_count": 1,
  "views": 0,
  "content": "<p>The host mentioned that there is some raw data available here: <a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202\" target=\"_blank\">https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202</a>.</p>\n<p>From reading other kagglers' comments, I guess there are more user sessions in the raw data  than in the training data we got, so I am wondering whether people have tried writing a parser for the raw data to incorporate those in the training. It would be great to hear your thoughts on this.</p>",
  "messages": [
    {
      "id": "2237509",
      "postDate": "04/27/2023 17:19:45",
      "content": "<p>The host mentioned that there is some raw data available here: <a href=\"https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202\" target=\"_blank\">https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202</a>.</p>\n<p>From reading other kagglers' comments, I guess there are more user sessions in the raw data  than in the training data we got, so I am wondering whether people have tried writing a parser for the raw data to incorporate those in the training. It would be great to hear your thoughts on this.</p>",
      "rawMarkdown": "The host mentioned that there is some raw data available here: https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202.\n\nFrom reading other kagglers' comments, I guess there are more user sessions in the raw data  than in the training data we got, so I am wondering whether people have tried writing a parser for the raw data to incorporate those in the training. It would be great to hear your thoughts on this.",
      "votes": null
    },
    {
      "id": "2237653",
      "postDate": "04/27/2023 20:20:34",
      "content": "<p>great question! I am curious to hear other people's opinion on that too..</p>",
      "rawMarkdown": "great question! I am curious to hear other people's opinion on that too..",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2237653,
      "author_name": "ngocuong",
      "author_url": "",
      "post_date": "04/27/2023 20:20:34",
      "content": "<p>great question! I am curious to hear other people's opinion on that too..</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2237509": "The host mentioned that there is some raw data available here: https://www.kaggle.com/competitions/predict-student-performance-from-game-play/discussion/396202.\n\nFrom reading other kagglers' comments, I guess there are more user sessions in the raw data  than in the training data we got, so I am wondering whether people have tried writing a parser for the raw data to incorporate those in the training. It would be great to hear your thoughts on this.",
    "2237653": "great question! I am curious to hear other people's opinion on that too.."
  },
  "source": "meta"
}