{
  "id": 391201,
  "title": "Kaggle, deal with this (please)",
  "url": "/competitions/rsna-breast-cancer-detection/discussion/391201",
  "author_name": "GUNER",
  "post_date": "2023-02-28T17:19:25.943000",
  "votes": 10,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Yes, the issue on external dataset. <br>\nThis really affects the results. </p>\n<p>Many people probably also within the first 100, did not use external dataset because they could not decide and there was no feedback. </p>\n<p>To be fair, some requests under different discussion threads have received an 'under topic' feedback as in the linked discussion, where the response came after weeks. [<a href=\"https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211\" target=\"_blank\">https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211</a>]</p>\n<p>Anyway, please do something, and it is not an unsolvable situation although might require different approaches in different competition. Here is an example waytogo which actually is built on external dataset rules from another Kaggle Competition (i.e. [<a href=\"https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414\" target=\"_blank\">https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414</a>]</p>\n<p><strong>1.</strong>    Make a seperate pinned discussion topic for external datasets for every competition.<br>\n<strong>2.</strong>    Make people bring their external dataset recommendations.<br>\n<strong>3.</strong>    Coordinate the grant/prohibit of each set to be used in the competition: this is mostly host issue, that is:  give the host a week for each brought up dataset, on the decision if it should be considered available or not.</p>\n<p>Bottomline: <br>\nEvery competition has its own hurdles and lives in a different environment, but the issue on external datasets is probably the most common weakest link in Kaggle Competition setup. <br>\nThis is better dealt with taking action rather than by taking inaction.</p>\n<p>I probably had to put this under 'Kaggle General' but it was also so relevant in this Competition.</p>\n<p>You can do this Kaggle! </p>",
  "messages": [
    {
      "id": 2163240,
      "postDate": "2023-02-28T17:19:25.943Z",
      "content": "<p>Yes, the issue on external dataset. <br>\nThis really affects the results. </p>\n<p>Many people probably also within the first 100, did not use external dataset because they could not decide and there was no feedback. </p>\n<p>To be fair, some requests under different discussion threads have received an 'under topic' feedback as in the linked discussion, where the response came after weeks. [<a href=\"https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211\" target=\"_blank\">https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211</a>]</p>\n<p>Anyway, please do something, and it is not an unsolvable situation although might require different approaches in different competition. Here is an example waytogo which actually is built on external dataset rules from another Kaggle Competition (i.e. [<a href=\"https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414\" target=\"_blank\">https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414</a>]</p>\n<p><strong>1.</strong>    Make a seperate pinned discussion topic for external datasets for every competition.<br>\n<strong>2.</strong>    Make people bring their external dataset recommendations.<br>\n<strong>3.</strong>    Coordinate the grant/prohibit of each set to be used in the competition: this is mostly host issue, that is:  give the host a week for each brought up dataset, on the decision if it should be considered available or not.</p>\n<p>Bottomline: <br>\nEvery competition has its own hurdles and lives in a different environment, but the issue on external datasets is probably the most common weakest link in Kaggle Competition setup. <br>\nThis is better dealt with taking action rather than by taking inaction.</p>\n<p>I probably had to put this under 'Kaggle General' but it was also so relevant in this Competition.</p>\n<p>You can do this Kaggle! </p>",
      "rawMarkdown": "Yes, the issue on external dataset. \nThis really affects the results. \n\nMany people probably also within the first 100, did not use external dataset because they could not decide and there was no feedback. \n\nTo be fair, some requests under different discussion threads have received an 'under topic' feedback as in the linked discussion, where the response came after weeks. [https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211]\n\nAnyway, please do something, and it is not an unsolvable situation although might require different approaches in different competition. Here is an example waytogo which actually is built on external dataset rules from another Kaggle Competition (i.e. [https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414]\n\n**1.**    Make a seperate pinned discussion topic for external datasets for every competition.\n**2.**    Make people bring their external dataset recommendations.\n**3.**    Coordinate the grant/prohibit of each set to be used in the competition: this is mostly host issue, that is:  give the host a week for each brought up dataset, on the decision if it should be considered available or not.\n\nBottomline: \nEvery competition has its own hurdles and lives in a different environment, but the issue on external datasets is probably the most common weakest link in Kaggle Competition setup. \nThis is better dealt with taking action rather than by taking inaction.\n\nI probably had to put this under 'Kaggle General' but it was also so relevant in this Competition.\n\nYou can do this Kaggle! \n\n\n",
      "votes": 10
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2163240": "Yes, the issue on external dataset. \nThis really affects the results. \n\nMany people probably also within the first 100, did not use external dataset because they could not decide and there was no feedback. \n\nTo be fair, some requests under different discussion threads have received an 'under topic' feedback as in the linked discussion, where the response came after weeks. [https://www.kaggle.com/competitions/rsna-breast-cancer-detection/discussion/373211]\n\nAnyway, please do something, and it is not an unsolvable situation although might require different approaches in different competition. Here is an example waytogo which actually is built on external dataset rules from another Kaggle Competition (i.e. [https://www.kaggle.com/competitions/coleridgeinitiative-show-us-the-data/discussion/232414]\n\n**1.**    Make a seperate pinned discussion topic for external datasets for every competition.\n**2.**    Make people bring their external dataset recommendations.\n**3.**    Coordinate the grant/prohibit of each set to be used in the competition: this is mostly host issue, that is:  give the host a week for each brought up dataset, on the decision if it should be considered available or not.\n\nBottomline: \nEvery competition has its own hurdles and lives in a different environment, but the issue on external datasets is probably the most common weakest link in Kaggle Competition setup. \nThis is better dealt with taking action rather than by taking inaction.\n\nI probably had to put this under 'Kaggle General' but it was also so relevant in this Competition.\n\nYou can do this Kaggle! \n\n\n"
  }
}