{
  "id": 6193,
  "title": "Explanation of the scores in leaderboard",
  "url": "/competitions/facebook-recruiting-iii-keyword-extraction/discussion/6193",
  "author_name": "",
  "post_date": "2013-11-01T18:02:31.307Z",
  "votes": null,
  "comment_count": 1,
  "views": 861,
  "content": "<p>Can somebody explain how the scores in the leaderboard are arrived at?</p>\n<p>What evaluation measures are they? F-measure? Precision@k, Recall?</p>\n<p>&nbsp;</p>",
  "messages": [
    {
      "id": "33044",
      "postDate": "11/01/2013 18:02:31",
      "content": "<p>Can somebody explain how the scores in the leaderboard are arrived at?</p>\n<p>What evaluation measures are they? F-measure? Precision@k, Recall?</p>\n<p>&nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "33048",
      "postDate": "11/01/2013 19:24:17",
      "content": "<p>https://www.kaggle.com/c/facebook-recruiting-iii-keyword-extraction/details/evaluation</p>\n<p>&nbsp;</p>\n<p>I'd suggest calculating F1 for a model which is the same as the benchmarks in the leaderboard so that you can verify that you are doing it correctly. You should get very similar numbers.</p>\n<p>&nbsp;</p>\n<p>This is useful for your implementation:</p>\n<p>http://en.wikipedia.org/wiki/Precision_and_recall#Definition_.28classification_context.29</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 33048,
      "author_name": "pminkov",
      "author_url": "",
      "post_date": "11/01/2013 19:24:17",
      "content": "<p>https://www.kaggle.com/c/facebook-recruiting-iii-keyword-extraction/details/evaluation</p>\n<p>&nbsp;</p>\n<p>I'd suggest calculating F1 for a model which is the same as the benchmarks in the leaderboard so that you can verify that you are doing it correctly. You should get very similar numbers.</p>\n<p>&nbsp;</p>\n<p>This is useful for your implementation:</p>\n<p>http://en.wikipedia.org/wiki/Precision_and_recall#Definition_.28classification_context.29</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "33044": "",
    "33048": ""
  },
  "source": "meta"
}