{
  "id": 5854,
  "title": "prop_location_score1 and prop_location_score2",
  "url": "/competitions/expedia-personalized-sort/discussion/5854",
  "author_name": "",
  "post_date": "2013-09-23T12:29:05.723Z",
  "votes": null,
  "comment_count": 3,
  "views": 2253,
  "content": "<p>How prop_location_score1 and prop_location_score2 scores are gathered?</p>\n<p>Is there any correlation between them?</p>\n<p>Is one more relevant than the other?</p>",
  "messages": [
    {
      "id": "31396",
      "postDate": "09/23/2013 12:29:05",
      "content": "<p>How prop_location_score1 and prop_location_score2 scores are gathered?</p>\n<p>Is there any correlation between them?</p>\n<p>Is one more relevant than the other?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "31397",
      "postDate": "09/23/2013 12:40:35",
      "content": "<p>1. I cannot reveal the details.</p>\n<p>2. They are positively correlated.</p>\n<p>3. It is for you to discover.</p>\n<p>&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "67621",
      "postDate": "03/22/2015 17:09:08",
      "content": "<p>From this and a couple of other posts, I understand that you cannot reveal how the location scores were determined and that to construct them you used several other features that are not included in the data set (but which were used to construct the position benchmark). I just wanted to make sure that even though these are some scores, they still measure the location of the hotel and how desirable it is for visitors. Is this right? So it's something like distance from center, proximity to a beach, but weighted by how desirable these are?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "69027",
      "postDate": "03/30/2015 15:30:52",
      "content": "<p>Yes, these factors measure desirability of a location. Distance from the center is used in the computation of those.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 31397,
      "author_name": "adamwoz",
      "author_url": "",
      "post_date": "09/23/2013 12:40:35",
      "content": "<p>1. I cannot reveal the details.</p>\n<p>2. They are positively correlated.</p>\n<p>3. It is for you to discover.</p>\n<p>&nbsp;</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 67621,
      "author_name": "ralucaursu",
      "author_url": "",
      "post_date": "03/22/2015 17:09:08",
      "content": "<p>From this and a couple of other posts, I understand that you cannot reveal how the location scores were determined and that to construct them you used several other features that are not included in the data set (but which were used to construct the position benchmark). I just wanted to make sure that even though these are some scores, they still measure the location of the hotel and how desirable it is for visitors. Is this right? So it's something like distance from center, proximity to a beach, but weighted by how desirable these are?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 69027,
      "author_name": "adamwoz",
      "author_url": "",
      "post_date": "03/30/2015 15:30:52",
      "content": "<p>Yes, these factors measure desirability of a location. Distance from the center is used in the computation of those.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "31396": "",
    "31397": "",
    "67621": "",
    "69027": ""
  },
  "source": "meta"
}