{
  "id": 193635,
  "title": "The question of 10033 doesn't have tags",
  "url": "/competitions/riiid-test-answer-prediction/discussion/193635",
  "author_name": "",
  "post_date": "2020-10-28T02:01:36.209576700Z",
  "votes": 3,
  "comment_count": 2,
  "views": 0,
  "content": "<p>The question of 10033 doesn't have tags.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1483555%2Ff81721800740773e54f58941dcab7fb5%2F2020-10-28%2010.56.56.png?generation=1603850413190475&amp;alt=media\" alt=\"\"></p>\n<p>My understanding, the question has one or more detailed tag codes.<br>\nHow to deal this question?</p>\n<p>I apologize if the same discussion has been made before. (I may have missed it.)</p>\n<p>Thanks.</p>",
  "messages": [
    {
      "id": "1062576",
      "postDate": "10/28/2020 02:01:36",
      "content": "<p>The question of 10033 doesn't have tags.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1483555%2Ff81721800740773e54f58941dcab7fb5%2F2020-10-28%2010.56.56.png?generation=1603850413190475&amp;alt=media\" alt=\"\"></p>\n<p>My understanding, the question has one or more detailed tag codes.<br>\nHow to deal this question?</p>\n<p>I apologize if the same discussion has been made before. (I may have missed it.)</p>\n<p>Thanks.</p>",
      "rawMarkdown": "The question of 10033 doesn't have tags.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1483555%2Ff81721800740773e54f58941dcab7fb5%2F2020-10-28%2010.56.56.png?generation=1603850413190475&alt=media)\n\nMy understanding, the question has one or more detailed tag codes.\nHow to deal this question?\n\nI apologize if the same discussion has been made before. (I may have missed it.)\n\nThanks.",
      "votes": null
    },
    {
      "id": "1062649",
      "postDate": "10/28/2020 04:21:20",
      "content": "<p>It’s just the one question (out of ~ 13.5K+) so it’s unlikely to make a difference in the model.</p>\n<p>Either consider it as having no tags or else impute it with the most common tag globally (or of the 6th part).</p>\n<p>From my experiments it won’t make a difference.</p>",
      "rawMarkdown": "It’s just the one question (out of ~ 13.5K+) so it’s unlikely to make a difference in the model.\n\nEither consider it as having no tags or else impute it with the most common tag globally (or of the 6th part).\n\nFrom my experiments it won’t make a difference.",
      "votes": null
    },
    {
      "id": "1062896",
      "postDate": "10/28/2020 09:40:39",
      "content": "<p>Thank you reply.</p>\n<p>I see.<br>\nI make question type embedding now, I treat this no tag.</p>",
      "rawMarkdown": "Thank you reply.\n\nI see.\nI make question type embedding now, I treat this no tag.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1062649,
      "author_name": "rohanrao",
      "author_url": "",
      "post_date": "10/28/2020 04:21:20",
      "content": "<p>It’s just the one question (out of ~ 13.5K+) so it’s unlikely to make a difference in the model.</p>\n<p>Either consider it as having no tags or else impute it with the most common tag globally (or of the 6th part).</p>\n<p>From my experiments it won’t make a difference.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1062896,
          "author_name": "takamichitoda",
          "author_url": "",
          "post_date": "10/28/2020 09:40:39",
          "content": "<p>Thank you reply.</p>\n<p>I see.<br>\nI make question type embedding now, I treat this no tag.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1062576": "The question of 10033 doesn't have tags.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1483555%2Ff81721800740773e54f58941dcab7fb5%2F2020-10-28%2010.56.56.png?generation=1603850413190475&alt=media)\n\nMy understanding, the question has one or more detailed tag codes.\nHow to deal this question?\n\nI apologize if the same discussion has been made before. (I may have missed it.)\n\nThanks.",
    "1062649": "It’s just the one question (out of ~ 13.5K+) so it’s unlikely to make a difference in the model.\n\nEither consider it as having no tags or else impute it with the most common tag globally (or of the 6th part).\n\nFrom my experiments it won’t make a difference.",
    "1062896": "Thank you reply.\n\nI see.\nI make question type embedding now, I treat this no tag."
  },
  "source": "meta"
}