{
  "id": 236753,
  "title": "About ADNI ",
  "url": "/competitions/coleridgeinitiative-show-us-the-data/discussion/236753",
  "author_name": "",
  "post_date": "2021-05-05T16:58:39.720952600Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Anyone understands when to predict just ADNI, when Alzheimer's Disease Neuroimaging Initative ADNI and when to predict Alzheimer's Disease Neuroimaging Initative ADNI | ADNI? It looks like this dataset has the largest number of spellings, besides ADNI and Alzheimer's Disease Neuroimaging Initative ADNI are essentially the same thing, do other datasets need to be processed this way, for example SLOSH and its full name?</p>",
  "messages": [
    {
      "id": "1294449",
      "postDate": "05/05/2021 16:58:39",
      "content": "<p>Anyone understands when to predict just ADNI, when Alzheimer's Disease Neuroimaging Initative ADNI and when to predict Alzheimer's Disease Neuroimaging Initative ADNI | ADNI? It looks like this dataset has the largest number of spellings, besides ADNI and Alzheimer's Disease Neuroimaging Initative ADNI are essentially the same thing, do other datasets need to be processed this way, for example SLOSH and its full name?</p>",
      "rawMarkdown": "Anyone understands when to predict just ADNI, when Alzheimer's Disease Neuroimaging Initative ADNI and when to predict Alzheimer's Disease Neuroimaging Initative ADNI | ADNI? It looks like this dataset has the largest number of spellings, besides ADNI and Alzheimer's Disease Neuroimaging Initative ADNI are essentially the same thing, do other datasets need to be processed this way, for example SLOSH and its full name?",
      "votes": null
    },
    {
      "id": "1294454",
      "postDate": "05/05/2021 17:00:20",
      "content": "<p>Hi Igor - I've asked a similar question to Julia in this discussion - <a href=\"url\" target=\"_blank\">https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236508</a></p>\n<p>Hopefully one of us will find the answer…</p>",
      "rawMarkdown": "Hi Igor - I've asked a similar question to Julia in this discussion - [https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236508](url)\n\nHopefully one of us will find the answer...",
      "votes": null
    },
    {
      "id": "1294510",
      "postDate": "05/05/2021 17:46:46",
      "content": "<p>Hello, Mark ! I found the answer <a href=\"https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/235297\" target=\"_blank\">here</a> The answer is \"Nobody know\" 😏</p>",
      "rawMarkdown": "Hello, Mark ! I found the answer [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/235297) The answer is \"Nobody know\" 😏",
      "votes": null
    },
    {
      "id": "1294523",
      "postDate": "05/05/2021 17:55:26",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/kucherevskiy\" target=\"_blank\">@kucherevskiy</a>  and <a href=\"https://www.kaggle.com/cascadenite\" target=\"_blank\">@cascadenite</a>  You are right that authors will cite the same dataset in different ways.    We've responded to that in the past (e.g. in the case of the data that I'm most familiar with, LEHD data is also called Longitudinal Employer Household Dynamics data).   Authors can also misspell data.   The challenge is to find how authors are referencing datasets in their research.   We're not challenging you to build a look up table.    I hope that helps</p>",
      "rawMarkdown": "Hi @kucherevskiy  and @cascadenite  You are right that authors will cite the same dataset in different ways.    We've responded to that in the past (e.g. in the case of the data that I'm most familiar with, LEHD data is also called Longitudinal Employer Household Dynamics data).   Authors can also misspell data.   The challenge is to find how authors are referencing datasets in their research.   We're not challenging you to build a look up table.    I hope that helps",
      "votes": null
    },
    {
      "id": "1295987",
      "postDate": "05/06/2021 21:22:29",
      "content": "<p><a href=\"https://www.kaggle.com/julialane\" target=\"_blank\">@julialane</a> Unfortunately it doesn't answer the question. The problem is that if a publication mentions only <strong>once</strong>:</p>\n<blockquote>\n  <p>Alzheimer's Disease Neuroimaging Initative ADNI</p>\n</blockquote>\n<p>we should expect the ground truth to be <code>Alzheimer's Disease Neuroimaging Initative ADNI</code>. Instead, from the public train set there are many cases in which the ground truth is both <code>ADNI</code> and <code>Alzheimer's Disease Neuroimaging Initative ADNI</code> as though the publication contains two distinct mentions. There seems to be very little consistency in the way the data are labelled so that the models cannot be evaluated in a reliable way. See <a href=\"https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972\" target=\"_blank\">here</a></p>",
      "rawMarkdown": "julialane Unfortunately it doesn't answer the question. The problem is that if a publication mentions only **once**:\n> Alzheimer's Disease Neuroimaging Initative ADNI\n\nwe should expect the ground truth to be `Alzheimer's Disease Neuroimaging Initative ADNI`. Instead, from the public train set there are many cases in which the ground truth is both `ADNI` and `Alzheimer's Disease Neuroimaging Initative ADNI` as though the publication contains two distinct mentions. There seems to be very little consistency in the way the data are labelled so that the models cannot be evaluated in a reliable way. See [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1294454,
      "author_name": "cascadenite",
      "author_url": "",
      "post_date": "05/05/2021 17:00:20",
      "content": "<p>Hi Igor - I've asked a similar question to Julia in this discussion - <a href=\"url\" target=\"_blank\">https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236508</a></p>\n<p>Hopefully one of us will find the answer…</p>",
      "votes": null,
      "replies": [
        {
          "id": 1294510,
          "author_name": "kucherevskiy",
          "author_url": "",
          "post_date": "05/05/2021 17:46:46",
          "content": "<p>Hello, Mark ! I found the answer <a href=\"https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/235297\" target=\"_blank\">here</a> The answer is \"Nobody know\" 😏</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1294523,
          "author_name": "julialane",
          "author_url": "",
          "post_date": "05/05/2021 17:55:26",
          "content": "<p>Hi <a href=\"https://www.kaggle.com/kucherevskiy\" target=\"_blank\">@kucherevskiy</a>  and <a href=\"https://www.kaggle.com/cascadenite\" target=\"_blank\">@cascadenite</a>  You are right that authors will cite the same dataset in different ways.    We've responded to that in the past (e.g. in the case of the data that I'm most familiar with, LEHD data is also called Longitudinal Employer Household Dynamics data).   Authors can also misspell data.   The challenge is to find how authors are referencing datasets in their research.   We're not challenging you to build a look up table.    I hope that helps</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1295987,
          "author_name": "eguidotti",
          "author_url": "",
          "post_date": "05/06/2021 21:22:29",
          "content": "<p><a href=\"https://www.kaggle.com/julialane\" target=\"_blank\">@julialane</a> Unfortunately it doesn't answer the question. The problem is that if a publication mentions only <strong>once</strong>:</p>\n<blockquote>\n  <p>Alzheimer's Disease Neuroimaging Initative ADNI</p>\n</blockquote>\n<p>we should expect the ground truth to be <code>Alzheimer's Disease Neuroimaging Initative ADNI</code>. Instead, from the public train set there are many cases in which the ground truth is both <code>ADNI</code> and <code>Alzheimer's Disease Neuroimaging Initative ADNI</code> as though the publication contains two distinct mentions. There seems to be very little consistency in the way the data are labelled so that the models cannot be evaluated in a reliable way. See <a href=\"https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972\" target=\"_blank\">here</a></p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1294449": "Anyone understands when to predict just ADNI, when Alzheimer's Disease Neuroimaging Initative ADNI and when to predict Alzheimer's Disease Neuroimaging Initative ADNI | ADNI? It looks like this dataset has the largest number of spellings, besides ADNI and Alzheimer's Disease Neuroimaging Initative ADNI are essentially the same thing, do other datasets need to be processed this way, for example SLOSH and its full name?",
    "1294454": "Hi Igor - I've asked a similar question to Julia in this discussion - [https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236508](url)\n\nHopefully one of us will find the answer...",
    "1294510": "Hello, Mark ! I found the answer [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/235297) The answer is \"Nobody know\" 😏",
    "1294523": "Hi @kucherevskiy  and @cascadenite  You are right that authors will cite the same dataset in different ways.    We've responded to that in the past (e.g. in the case of the data that I'm most familiar with, LEHD data is also called Longitudinal Employer Household Dynamics data).   Authors can also misspell data.   The challenge is to find how authors are referencing datasets in their research.   We're not challenging you to build a look up table.    I hope that helps",
    "1295987": "julialane Unfortunately it doesn't answer the question. The problem is that if a publication mentions only **once**:\n> Alzheimer's Disease Neuroimaging Initative ADNI\n\nwe should expect the ground truth to be `Alzheimer's Disease Neuroimaging Initative ADNI`. Instead, from the public train set there are many cases in which the ground truth is both `ADNI` and `Alzheimer's Disease Neuroimaging Initative ADNI` as though the publication contains two distinct mentions. There seems to be very little consistency in the way the data are labelled so that the models cannot be evaluated in a reliable way. See [here](https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/discussion/236972)"
  },
  "source": "meta"
}