{
  "id": 24186,
  "title": "New to competition",
  "url": "/competitions/outbrain-click-prediction/discussion/24186",
  "author_name": "",
  "post_date": "2016-10-08T02:57:49.017Z",
  "votes": null,
  "comment_count": 3,
  "views": 483,
  "content": "<p>Hello everyone,\n                I am new to data science. I have checked the data and found that following files have one column named &quot;Confidence&quot;:\n  1. documents_categories\n  2. documents_topics</p>\n\n<p>What do they mean? \nI was assuming that they mean probabilities of selecting topic or document category. Is I my understanding correct? </p>",
  "messages": [
    {
      "id": "138326",
      "postDate": "10/08/2016 02:57:49",
      "content": "<p>Hello everyone,\n                I am new to data science. I have checked the data and found that following files have one column named &quot;Confidence&quot;:\n  1. documents_categories\n  2. documents_topics</p>\n\n<p>What do they mean? \nI was assuming that they mean probabilities of selecting topic or document category. Is I my understanding correct? </p>",
      "rawMarkdown": "Hello everyone,\r\n                I am new to data science. I have checked the data and found that following files have one column named \"Confidence\":\r\n  1. documents_categories\r\n  2. documents_topics\r\n  \r\nWhat do they mean? \r\nI was assuming that they mean probabilities of selecting topic or document category. Is I my understanding correct?",
      "votes": null
    },
    {
      "id": "138359",
      "postDate": "10/08/2016 08:02:24",
      "content": "<p>Hi Srini. From the 'Data' page of this competition:</p>\n\n<blockquote>\n  <p>documents_topics.csv, documents_entities.csv, and documents_categories.csv all provide information about the content in a document, as well as Outbrain's confidence in each respective relationship. For example, an entity_id can represent a person, organization, or location. The rows in documents_entities.csv give the confidence that the given entity was referred to in the document.</p>\n</blockquote>\n\n<p>So Outbrain has done some internal classification and is providing the confidence of those classifications.</p>",
      "rawMarkdown": "Hi Srini. From the 'Data' page of this competition:\r\n\r\n> documents_topics.csv, documents_entities.csv, and documents_categories.csv all provide information about the content in a document, as well as Outbrain's confidence in each respective relationship. For example, an entity_id can represent a person, organization, or location. The rows in documents_entities.csv give the confidence that the given entity was referred to in the document.\r\n\r\nSo Outbrain has done some internal classification and is providing the confidence of those classifications.",
      "votes": null
    },
    {
      "id": "138369",
      "postDate": "10/08/2016 09:47:58",
      "content": "<p>When you are saying confidence of those classification, it means class probabilities right??</p>",
      "rawMarkdown": "When you are saying confidence of those classification, it means class probabilities right??",
      "votes": null
    },
    {
      "id": "138374",
      "postDate": "10/08/2016 10:45:21",
      "content": "<p>Quite probably, with the usual caveat that not all ML algorithms estimate this well.</p>",
      "rawMarkdown": "Quite probably, with the usual caveat that not all ML algorithms estimate this well.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 138359,
      "author_name": "joconnor",
      "author_url": "",
      "post_date": "10/08/2016 08:02:24",
      "content": "<p>Hi Srini. From the 'Data' page of this competition:</p>\n\n<blockquote>\n  <p>documents_topics.csv, documents_entities.csv, and documents_categories.csv all provide information about the content in a document, as well as Outbrain's confidence in each respective relationship. For example, an entity_id can represent a person, organization, or location. The rows in documents_entities.csv give the confidence that the given entity was referred to in the document.</p>\n</blockquote>\n\n<p>So Outbrain has done some internal classification and is providing the confidence of those classifications.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 138369,
      "author_name": "sr71nighthawk",
      "author_url": "",
      "post_date": "10/08/2016 09:47:58",
      "content": "<p>When you are saying confidence of those classification, it means class probabilities right??</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 138374,
      "author_name": "joconnor",
      "author_url": "",
      "post_date": "10/08/2016 10:45:21",
      "content": "<p>Quite probably, with the usual caveat that not all ML algorithms estimate this well.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "138326": "Hello everyone,\r\n                I am new to data science. I have checked the data and found that following files have one column named \"Confidence\":\r\n  1. documents_categories\r\n  2. documents_topics\r\n  \r\nWhat do they mean? \r\nI was assuming that they mean probabilities of selecting topic or document category. Is I my understanding correct?",
    "138359": "Hi Srini. From the 'Data' page of this competition:\r\n\r\n> documents_topics.csv, documents_entities.csv, and documents_categories.csv all provide information about the content in a document, as well as Outbrain's confidence in each respective relationship. For example, an entity_id can represent a person, organization, or location. The rows in documents_entities.csv give the confidence that the given entity was referred to in the document.\r\n\r\nSo Outbrain has done some internal classification and is providing the confidence of those classifications.",
    "138369": "When you are saying confidence of those classification, it means class probabilities right??",
    "138374": "Quite probably, with the usual caveat that not all ML algorithms estimate this well."
  },
  "source": "meta"
}