{
  "id": 20875,
  "title": "Questions about the data",
  "url": "/competitions/facebook-v-predicting-check-ins/discussion/20875",
  "author_name": "",
  "post_date": "2016-05-11T22:50:14.393Z",
  "votes": null,
  "comment_count": 6,
  "views": 1657,
  "content": "<p>/summon @Anaconda &amp; @Ampersand:</p>\n\n<ul>\n<li><p>What is the meaning of the accuracy values? Does a high number implies a high accuracy? Or a low accuracy? </p></li>\n<li><p>Are the places static? (i.e. for a given place_id, does x and y stay unchanged?) Or do we have mobile locations?</p></li>\n<li><p>What is the time unit used for timestamps? Seconds, milliseconds?</p></li>\n</ul>\n\n<p>Thanks!</p>",
  "messages": [
    {
      "id": "119623",
      "postDate": "05/11/2016 22:50:14",
      "content": "<p>/summon @Anaconda &amp; @Ampersand:</p>\n\n<ul>\n<li><p>What is the meaning of the accuracy values? Does a high number implies a high accuracy? Or a low accuracy? </p></li>\n<li><p>Are the places static? (i.e. for a given place_id, does x and y stay unchanged?) Or do we have mobile locations?</p></li>\n<li><p>What is the time unit used for timestamps? Seconds, milliseconds?</p></li>\n</ul>\n\n<p>Thanks!</p>",
      "rawMarkdown": "/summon @Anaconda & @Ampersand:\r\n\r\n* What is the meaning of the accuracy values? Does a high number implies a high accuracy? Or a low accuracy? \r\n\r\n* Are the places static? (i.e. for a given place_id, does x and y stay unchanged?) Or do we have mobile locations?\r\n\r\n* What is the time unit used for timestamps? Seconds, milliseconds?\r\n\r\nThanks!",
      "votes": null
    },
    {
      "id": "119624",
      "postDate": "05/11/2016 22:56:00",
      "content": "<p>From the data page <a href=\"https://www.kaggle.com/c/facebook-v-predicting-check-ins/data\">link</a>:\nNote: Some of the columns, such as time and accuracy, are intentionally left vague in their definitions. Please consider them as part of the challenge. Personally my guess is accuracy is some measure of how reliable the x and y position measurements are.</p>",
      "rawMarkdown": "From the data page [link](https://www.kaggle.com/c/facebook-v-predicting-check-ins/data):\r\nNote: Some of the columns, such as time and accuracy, are intentionally left vague in their definitions. Please consider them as part of the challenge. Personally my guess is accuracy is some measure of how reliable the x and y position measurements are.",
      "votes": null
    },
    {
      "id": "119642",
      "postDate": "05/12/2016 01:46:40",
      "content": "<ol>\n<li>Yes, we don't have much more to add apart from what is written in the data page. While both accuracy &amp; time, to some extent, resemble reality, they were fabricated specifically for the purpose of this competition. The best source of insight is the data itself.</li>\n<li>&quot;Re: for a given place_id, does x and y stay unchanged?&quot;: [x,y] coordinates represent check-in coordinates, not place coordinates.\nRepresenting places as points may be a good first proxy, but real\nplaces have various shapes. Also, note that people may check into\ndifferent places even though they are not physically there.</li>\n</ol>",
      "rawMarkdown": "1. Yes, we don't have much more to add apart from what is written in the data page. While both accuracy & time, to some extent, resemble reality, they were fabricated specifically for the purpose of this competition. The best source of insight is the data itself.\r\n 2. \"Re: for a given place_id, does x and y stay unchanged?\": [x,y] coordinates represent check-in coordinates, not place coordinates.\r\n    Representing places as points may be a good first proxy, but real\r\n    places have various shapes. Also, note that people may check into\r\n    different places even though they are not physically there.",
      "votes": null
    },
    {
      "id": "119702",
      "postDate": "05/12/2016 11:27:50",
      "content": "<p>Awesome. Thanks for your answers!</p>",
      "rawMarkdown": "Awesome. Thanks for your answers!",
      "votes": null
    },
    {
      "id": "119720",
      "postDate": "05/12/2016 13:30:06",
      "content": "<p>@anaconda: Regarding your point on place_id, i have a question. While we are using Facebook and want to check-in to some place, even though it's not necessary to be at the place to check-in we need to be in a range of that place to be able to even see the option to check-in. Hence can we assume this x,y coordinate is &quot;nearby&quot; the place_id? Please let me know</p>",
      "rawMarkdown": "anaconda: Regarding your point on place_id, i have a question. While we are using Facebook and want to check-in to some place, even though it's not necessary to be at the place to check-in we need to be in a range of that place to be able to even see the option to check-in. Hence can we assume this x,y coordinate is \"nearby\" the place_id? Please let me know",
      "votes": null
    },
    {
      "id": "119723",
      "postDate": "05/12/2016 13:44:29",
      "content": "<p>@Advanced Research and Analytics: I believe your hypothesis could be verified from the training data; while FB users can check in to any place regardless of their distance to it, they are certainly more likely to check-in when they are &quot;nearby&quot;</p>",
      "rawMarkdown": "Advanced Research and Analytics: I believe your hypothesis could be verified from the training data; while FB users can check in to any place regardless of their distance to it, they are certainly more likely to check-in when they are \"nearby\"",
      "votes": null
    },
    {
      "id": "119738",
      "postDate": "05/12/2016 14:42:53",
      "content": "<p>Understood. Thanks</p>",
      "rawMarkdown": "Understood. Thanks",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 119624,
      "author_name": "jturkewitz",
      "author_url": "",
      "post_date": "05/11/2016 22:56:00",
      "content": "<p>From the data page <a href=\"https://www.kaggle.com/c/facebook-v-predicting-check-ins/data\">link</a>:\nNote: Some of the columns, such as time and accuracy, are intentionally left vague in their definitions. Please consider them as part of the challenge. Personally my guess is accuracy is some measure of how reliable the x and y position measurements are.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 119642,
      "author_name": "anaconda",
      "author_url": "",
      "post_date": "05/12/2016 01:46:40",
      "content": "<ol>\n<li>Yes, we don't have much more to add apart from what is written in the data page. While both accuracy &amp; time, to some extent, resemble reality, they were fabricated specifically for the purpose of this competition. The best source of insight is the data itself.</li>\n<li>&quot;Re: for a given place_id, does x and y stay unchanged?&quot;: [x,y] coordinates represent check-in coordinates, not place coordinates.\nRepresenting places as points may be a good first proxy, but real\nplaces have various shapes. Also, note that people may check into\ndifferent places even though they are not physically there.</li>\n</ol>",
      "votes": null,
      "replies": []
    },
    {
      "id": 119702,
      "author_name": "carrdelling",
      "author_url": "",
      "post_date": "05/12/2016 11:27:50",
      "content": "<p>Awesome. Thanks for your answers!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 119720,
      "author_name": "rama2810",
      "author_url": "",
      "post_date": "05/12/2016 13:30:06",
      "content": "<p>@anaconda: Regarding your point on place_id, i have a question. While we are using Facebook and want to check-in to some place, even though it's not necessary to be at the place to check-in we need to be in a range of that place to be able to even see the option to check-in. Hence can we assume this x,y coordinate is &quot;nearby&quot; the place_id? Please let me know</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 119723,
      "author_name": "anaconda",
      "author_url": "",
      "post_date": "05/12/2016 13:44:29",
      "content": "<p>@Advanced Research and Analytics: I believe your hypothesis could be verified from the training data; while FB users can check in to any place regardless of their distance to it, they are certainly more likely to check-in when they are &quot;nearby&quot;</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 119738,
      "author_name": "rama2810",
      "author_url": "",
      "post_date": "05/12/2016 14:42:53",
      "content": "<p>Understood. Thanks</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "119623": "/summon @Anaconda & @Ampersand:\r\n\r\n* What is the meaning of the accuracy values? Does a high number implies a high accuracy? Or a low accuracy? \r\n\r\n* Are the places static? (i.e. for a given place_id, does x and y stay unchanged?) Or do we have mobile locations?\r\n\r\n* What is the time unit used for timestamps? Seconds, milliseconds?\r\n\r\nThanks!",
    "119624": "From the data page [link](https://www.kaggle.com/c/facebook-v-predicting-check-ins/data):\r\nNote: Some of the columns, such as time and accuracy, are intentionally left vague in their definitions. Please consider them as part of the challenge. Personally my guess is accuracy is some measure of how reliable the x and y position measurements are.",
    "119642": "1. Yes, we don't have much more to add apart from what is written in the data page. While both accuracy & time, to some extent, resemble reality, they were fabricated specifically for the purpose of this competition. The best source of insight is the data itself.\r\n 2. \"Re: for a given place_id, does x and y stay unchanged?\": [x,y] coordinates represent check-in coordinates, not place coordinates.\r\n    Representing places as points may be a good first proxy, but real\r\n    places have various shapes. Also, note that people may check into\r\n    different places even though they are not physically there.",
    "119702": "Awesome. Thanks for your answers!",
    "119720": "anaconda: Regarding your point on place_id, i have a question. While we are using Facebook and want to check-in to some place, even though it's not necessary to be at the place to check-in we need to be in a range of that place to be able to even see the option to check-in. Hence can we assume this x,y coordinate is \"nearby\" the place_id? Please let me know",
    "119723": "Advanced Research and Analytics: I believe your hypothesis could be verified from the training data; while FB users can check in to any place regardless of their distance to it, they are certainly more likely to check-in when they are \"nearby\"",
    "119738": "Understood. Thanks"
  },
  "source": "meta"
}