{
  "id": 171051,
  "title": "Understanding dataset folder structure",
  "url": "/competitions/landmark-recognition-2020/discussion/171051",
  "author_name": "",
  "post_date": "2020-07-30T07:25:32.629860Z",
  "votes": 5,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Since there are a large number of images, each image is placed within three subfolders according to the first three characters of the image id (i.e. image abcdef.jpg is placed in a/b/c/abcdef.jpg).</p>",
  "messages": [
    {
      "id": "951508",
      "postDate": "07/30/2020 07:25:32",
      "content": "<p>Since there are a large number of images, each image is placed within three subfolders according to the first three characters of the image id (i.e. image abcdef.jpg is placed in a/b/c/abcdef.jpg).</p>",
      "rawMarkdown": "Since there are a large number of images, each image is placed within three subfolders according to the first three characters of the image id (i.e. image abcdef.jpg is placed in a/b/c/abcdef.jpg).",
      "votes": null
    },
    {
      "id": "962149",
      "postDate": "08/07/2020 21:07:53",
      "content": "<p>Hey, please could you explain to me what the second column of the sample submission csv file means. I don't understand what the numbers represent</p>",
      "rawMarkdown": "Hey, please could you explain to me what the second column of the sample submission csv file means. I don't understand what the numbers represent",
      "votes": null
    },
    {
      "id": "962752",
      "postDate": "08/08/2020 12:07:42",
      "content": "<p>The second column contains landmark predictions and must be either empty or contain exactly two numbers:\n- <em>landmark</em> (or *class_id* in terms of classification) which represent which landmark you assigned to this image\n- <em>confidence</em> score - how confident algorithm about assigning this landmark to image </p>",
      "rawMarkdown": "The second column contains landmark predictions and must be either empty or contain exactly two numbers:\n- *landmark* (or *class_id* in terms of classification) which represent which landmark you assigned to this image\n- *confidence* score - how confident algorithm about assigning this landmark to image",
      "votes": null
    },
    {
      "id": "963671",
      "postDate": "08/09/2020 07:46:07",
      "content": "<p>ooo wow. Thanks a lot that was really helpful</p>",
      "rawMarkdown": "ooo wow. Thanks a lot that was really helpful",
      "votes": null
    },
    {
      "id": "964167",
      "postDate": "08/09/2020 16:39:41",
      "content": "<p>Another thing to notice: as far as my exploration has gone, images of a given folder aren't related: for instance, have a look at the 0/0/0 train folder and you will see a variety of themes.</p>",
      "rawMarkdown": "Another thing to notice: as far as my exploration has gone, images of a given folder aren't related: for instance, have a look at the 0/0/0 train folder and you will see a variety of themes.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 962149,
      "author_name": "jonathanmbatuegwu",
      "author_url": "",
      "post_date": "08/07/2020 21:07:53",
      "content": "<p>Hey, please could you explain to me what the second column of the sample submission csv file means. I don't understand what the numbers represent</p>",
      "votes": null,
      "replies": [
        {
          "id": 962752,
          "author_name": "alexkirnas",
          "author_url": "",
          "post_date": "08/08/2020 12:07:42",
          "content": "<p>The second column contains landmark predictions and must be either empty or contain exactly two numbers:\n- <em>landmark</em> (or *class_id* in terms of classification) which represent which landmark you assigned to this image\n- <em>confidence</em> score - how confident algorithm about assigning this landmark to image </p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 963671,
          "author_name": "jonathanmbatuegwu",
          "author_url": "",
          "post_date": "08/09/2020 07:46:07",
          "content": "<p>ooo wow. Thanks a lot that was really helpful</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 964167,
      "author_name": "yassinealouini",
      "author_url": "",
      "post_date": "08/09/2020 16:39:41",
      "content": "<p>Another thing to notice: as far as my exploration has gone, images of a given folder aren't related: for instance, have a look at the 0/0/0 train folder and you will see a variety of themes.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "951508": "Since there are a large number of images, each image is placed within three subfolders according to the first three characters of the image id (i.e. image abcdef.jpg is placed in a/b/c/abcdef.jpg).",
    "962149": "Hey, please could you explain to me what the second column of the sample submission csv file means. I don't understand what the numbers represent",
    "962752": "The second column contains landmark predictions and must be either empty or contain exactly two numbers:\n- *landmark* (or *class_id* in terms of classification) which represent which landmark you assigned to this image\n- *confidence* score - how confident algorithm about assigning this landmark to image",
    "963671": "ooo wow. Thanks a lot that was really helpful",
    "964167": "Another thing to notice: as far as my exploration has gone, images of a given folder aren't related: for instance, have a look at the 0/0/0 train folder and you will see a variety of themes."
  },
  "source": "meta"
}