{
  "id": 585820,
  "title": "More data?",
  "url": "/competitions/aeroclub-recsys-2025/discussion/585820",
  "author_name": "",
  "post_date": "2025-06-23T09:57:05.892951900Z",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Dear fellow Kagglers, it's been almost a week since the competition launched, and I'm excited to see the momentum building… Despite organizers team initial concerns about training data size and Kaggle notebooks memory constraints, you're doing amazing work. </p>\n<p>The current dataset strikes a balance between being informative and manageable in size. However, we actually have significantly more data at our disposal. We're open to releasing additional training data, but we'd love to hear from you first:</p>\n<p><strong>What specific data would help you push your models to the next level given the current task setup?</strong></p>\n<p>Looking forward to your thoughts! 🤓</p>",
  "messages": [
    {
      "id": "3230660",
      "postDate": "06/23/2025 09:57:05",
      "content": "<p>Dear fellow Kagglers, it's been almost a week since the competition launched, and I'm excited to see the momentum building… Despite organizers team initial concerns about training data size and Kaggle notebooks memory constraints, you're doing amazing work. </p>\n<p>The current dataset strikes a balance between being informative and manageable in size. However, we actually have significantly more data at our disposal. We're open to releasing additional training data, but we'd love to hear from you first:</p>\n<p><strong>What specific data would help you push your models to the next level given the current task setup?</strong></p>\n<p>Looking forward to your thoughts! 🤓</p>",
      "rawMarkdown": "Dear fellow Kagglers, it's been almost a week since the competition launched, and I'm excited to see the momentum building... Despite organizers team initial concerns about training data size and Kaggle notebooks memory constraints, you're doing amazing work. \n\nThe current dataset strikes a balance between being informative and manageable in size. However, we actually have significantly more data at our disposal. We're open to releasing additional training data, but we'd love to hear from you first:\n\n**What specific data would help you push your models to the next level given the current task setup?**\n\nLooking forward to your thoughts! 🤓",
      "votes": null
    },
    {
      "id": "3231134",
      "postDate": "06/24/2025 00:55:43",
      "content": "<p>Thank you for your support! I believe it could be particularly helpful to include data from November–December 2023, if available. This would allow us to check for potential seasonal or year-end patterns, which might be relevant if the task exhibits any form of <strong>annual</strong> cyclicity.</p>",
      "rawMarkdown": "Thank you for your support! I believe it could be particularly helpful to include data from November–December 2023, if available. This would allow us to check for potential seasonal or year-end patterns, which might be relevant if the task exhibits any form of **annual** cyclicity.",
      "votes": null
    },
    {
      "id": "3235016",
      "postDate": "06/28/2025 15:51:16",
      "content": "<p>I think that if you divide the data into history and training, for calculating counters by target, then more data can help, but if you don't use counters, then this data is enough, as it seems to me)</p>",
      "rawMarkdown": "I think that if you divide the data into history and training, for calculating counters by target, then more data can help, but if you don't use counters, then this data is enough, as it seems to me)",
      "votes": null
    },
    {
      "id": "3235452",
      "postDate": "06/29/2025 09:13:41",
      "content": "<p>what other data do you have?</p>",
      "rawMarkdown": "what other data do you have?",
      "votes": null
    },
    {
      "id": "3238576",
      "postDate": "07/02/2025 02:04:47",
      "content": "<p>User location, what type of seats are available</p>",
      "rawMarkdown": "User location, what type of seats are available",
      "votes": null
    },
    {
      "id": "3239167",
      "postDate": "07/02/2025 14:55:02",
      "content": "<p>1.Agree on user location, 2 something about company - may be business type or so, location, people count.  ( history of cause good too, but size of df….   )</p>",
      "rawMarkdown": "1.Agree on user location, 2 something about company - may be business type or so, location, people count.  ( history of cause good too, but size of df....   )",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3231134,
      "author_name": "ka1242",
      "author_url": "",
      "post_date": "06/24/2025 00:55:43",
      "content": "<p>Thank you for your support! I believe it could be particularly helpful to include data from November–December 2023, if available. This would allow us to check for potential seasonal or year-end patterns, which might be relevant if the task exhibits any form of <strong>annual</strong> cyclicity.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3235016,
      "author_name": "insuperabile",
      "author_url": "",
      "post_date": "06/28/2025 15:51:16",
      "content": "<p>I think that if you divide the data into history and training, for calculating counters by target, then more data can help, but if you don't use counters, then this data is enough, as it seems to me)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3235452,
      "author_name": "junhaochan",
      "author_url": "",
      "post_date": "06/29/2025 09:13:41",
      "content": "<p>what other data do you have?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3238576,
      "author_name": "elpres1dente",
      "author_url": "",
      "post_date": "07/02/2025 02:04:47",
      "content": "<p>User location, what type of seats are available</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3239167,
      "author_name": "sergeyqt2024",
      "author_url": "",
      "post_date": "07/02/2025 14:55:02",
      "content": "<p>1.Agree on user location, 2 something about company - may be business type or so, location, people count.  ( history of cause good too, but size of df….   )</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3230660": "Dear fellow Kagglers, it's been almost a week since the competition launched, and I'm excited to see the momentum building... Despite organizers team initial concerns about training data size and Kaggle notebooks memory constraints, you're doing amazing work. \n\nThe current dataset strikes a balance between being informative and manageable in size. However, we actually have significantly more data at our disposal. We're open to releasing additional training data, but we'd love to hear from you first:\n\n**What specific data would help you push your models to the next level given the current task setup?**\n\nLooking forward to your thoughts! 🤓",
    "3231134": "Thank you for your support! I believe it could be particularly helpful to include data from November–December 2023, if available. This would allow us to check for potential seasonal or year-end patterns, which might be relevant if the task exhibits any form of **annual** cyclicity.",
    "3235016": "I think that if you divide the data into history and training, for calculating counters by target, then more data can help, but if you don't use counters, then this data is enough, as it seems to me)",
    "3235452": "what other data do you have?",
    "3238576": "User location, what type of seats are available",
    "3239167": "1.Agree on user location, 2 something about company - may be business type or so, location, people count.  ( history of cause good too, but size of df....   )"
  },
  "source": "meta"
}