{
  "id": 24274,
  "title": "confidence level",
  "url": "/competitions/outbrain-click-prediction/discussion/24274",
  "author_name": "",
  "post_date": "2016-10-10T23:59:44.813Z",
  "votes": 2,
  "comment_count": 3,
  "views": 278,
  "content": "<p>Cam anyone please explain what does confidence level mean in document_entities, document_category, document_ topics.\nOur team first thought that it could be the weight used to adjust the data, but it turns out that they cannot be sum to one.\nThanks!</p>",
  "messages": [
    {
      "id": "138809",
      "postDate": "10/10/2016 23:59:44",
      "content": "<p>Cam anyone please explain what does confidence level mean in document_entities, document_category, document_ topics.\nOur team first thought that it could be the weight used to adjust the data, but it turns out that they cannot be sum to one.\nThanks!</p>",
      "rawMarkdown": "Cam anyone please explain what does confidence level mean in document_entities, document_category, document_ topics.\r\nOur team first thought that it could be the weight used to adjust the data, but it turns out that they cannot be sum to one.\r\nThanks!",
      "votes": null
    },
    {
      "id": "138817",
      "postDate": "10/11/2016 01:03:54",
      "content": "<p>As I understand, it's a confidence level that the particular document really belongs to the particular category. This uncertainty arises because this information is most likely come from web-parsing and not reliable at 100%.</p>",
      "rawMarkdown": "As I understand, it's a confidence level that the particular document really belongs to the particular category. This uncertainty arises because this information is most likely come from web-parsing and not reliable at 100%.",
      "votes": null
    },
    {
      "id": "176695",
      "postDate": "04/21/2017 14:54:37",
      "content": "<p>Uncertainlty is fine. But what does the number itself signify? On what scale is it measured? If I measure it from  1-100 and if the evaluation program (not for this competition, I am working on the yt8m) uses a 0-5 scale, how could it ever match?</p>",
      "rawMarkdown": "Uncertainlty is fine. But what does the number itself signify? On what scale is it measured? If I measure it from  1-100 and if the evaluation program (not for this competition, I am working on the yt8m) uses a 0-5 scale, how could it ever match?",
      "votes": null
    },
    {
      "id": "232068",
      "postDate": "10/16/2017 18:26:00",
      "content": "<p>Hey guys I want to somehow extract ad headlines and topics in text is this possible with this dataset?</p>",
      "rawMarkdown": "Hey guys I want to somehow extract ad headlines and topics in text is this possible with this dataset?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 138817,
      "author_name": "santyago",
      "author_url": "",
      "post_date": "10/11/2016 01:03:54",
      "content": "<p>As I understand, it's a confidence level that the particular document really belongs to the particular category. This uncertainty arises because this information is most likely come from web-parsing and not reliable at 100%.</p>",
      "votes": null,
      "replies": [
        {
          "id": 176695,
          "author_name": "kas123",
          "author_url": "",
          "post_date": "04/21/2017 14:54:37",
          "content": "<p>Uncertainlty is fine. But what does the number itself signify? On what scale is it measured? If I measure it from  1-100 and if the evaluation program (not for this competition, I am working on the yt8m) uses a 0-5 scale, how could it ever match?</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 232068,
      "author_name": "clickstream",
      "author_url": "",
      "post_date": "10/16/2017 18:26:00",
      "content": "<p>Hey guys I want to somehow extract ad headlines and topics in text is this possible with this dataset?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "138809": "Cam anyone please explain what does confidence level mean in document_entities, document_category, document_ topics.\r\nOur team first thought that it could be the weight used to adjust the data, but it turns out that they cannot be sum to one.\r\nThanks!",
    "138817": "As I understand, it's a confidence level that the particular document really belongs to the particular category. This uncertainty arises because this information is most likely come from web-parsing and not reliable at 100%.",
    "176695": "Uncertainlty is fine. But what does the number itself signify? On what scale is it measured? If I measure it from  1-100 and if the evaluation program (not for this competition, I am working on the yt8m) uses a 0-5 scale, how could it ever match?",
    "232068": "Hey guys I want to somehow extract ad headlines and topics in text is this possible with this dataset?"
  },
  "source": "meta"
}