December 7, 2022 Dataset Open Access

Hiba Arnaout; Simon Razniewski; Jeff Z. Pan

JSON Export

{
  "files": [
    {
      "links": {
        "self": "https://zenodo.org/api/files/4114f475-9bed-403f-93f1-a03bd3fd7782/group_centric.jsonl"
      }, 
      "checksum": "md5:7143c37a9190cdc2628ca64266b0fc99", 
      "bucket": "4114f475-9bed-403f-93f1-a03bd3fd7782", 
      "key": "group_centric.jsonl", 
      "type": "jsonl", 
      "size": 63717789
    }, 
    {
      "links": {
        "self": "https://zenodo.org/api/files/4114f475-9bed-403f-93f1-a03bd3fd7782/subject_centric.jsonl"
      }, 
      "checksum": "md5:578265b26d600f6ecb1bccf7dcfe2438", 
      "bucket": "4114f475-9bed-403f-93f1-a03bd3fd7782", 
      "key": "subject_centric.jsonl", 
      "type": "jsonl", 
      "size": 172036359
    }
  ], 
  "owners": [
    455555
  ], 
  "doi": "10.5281/zenodo.7537200", 
  "stats": {
    "version_unique_downloads": 11.0, 
    "unique_views": 47.0, 
    "views": 53.0, 
    "version_views": 110.0, 
    "unique_downloads": 5.0, 
    "version_unique_views": 80.0, 
    "volume": 834698022.0, 
    "version_downloads": 15.0, 
    "downloads": 8.0, 
    "version_volume": 1209805519.0
  }, 
  "links": {
    "doi": "https://doi.org/10.5281/zenodo.7537200", 
    "conceptdoi": "https://doi.org/10.5281/zenodo.7410436", 
    "bucket": "https://zenodo.org/api/files/4114f475-9bed-403f-93f1-a03bd3fd7782", 
    "conceptbadge": "https://zenodo.org/badge/doi/10.5281/zenodo.7410436.svg", 
    "html": "https://zenodo.org/record/7537200", 
    "latest_html": "https://zenodo.org/record/7537200", 
    "badge": "https://zenodo.org/badge/doi/10.5281/zenodo.7537200.svg", 
    "latest": "https://zenodo.org/api/records/7537200"
  }, 
  "conceptdoi": "10.5281/zenodo.7410436", 
  "created": "2023-01-14T15:42:06.854775+00:00", 
  "updated": "2023-01-15T14:26:39.518770+00:00", 
  "conceptrecid": "7410436", 
  "revision": 3, 
  "id": 7537200, 
  "metadata": {
    "access_right_category": "success", 
    "doi": "10.5281/zenodo.7537200", 
    "description": "<p>These datasets contains&nbsp;statements about demographics and outliers&nbsp;of Wiki-based Communities of Interest.&nbsp;</p>\n\n<p><strong>Group-centric dataset (sample):</strong></p>\n\n<pre><code class=\"language-json\">{\n\t\"title\": \"winners of Priestley Medal\", \n\t\"recorded_members\": 83, \n\t\"topics\": [\"STEM.Chemistry\"], \n\t\"demographics\": [\n            \"occupation-chemist\",\n            \"gender-male\", \n            \"citizen-U.S.\"\n\t], \n\t\"outliers\": [\n\t\t{\n\t\t\t\"reason\": \"NOT(chemist) unlike 82 recorded members\", \n\t\t\t\"members\": [\n            \"Francis Garvan (lawyer, art collector)\"\n            ]\n\t\t}, \n\t\t{\n\t\t\t\"reason\": \"NOT(male) unlike 80 recorded members\", \n\t\t\t\"members\": [\n            \"Mary L. Good (female)\",\n            \"Darleane Hoffman (female)\", \n            \"Jacqueline Barton (female)\"\n            ]\n\t\t}\n\t]\n}</code></pre>\n\n<p><strong>Subject-centric dataset (sample):</strong></p>\n\n<pre><code class=\"language-json\">{\n\t\"subject\": \"Serena Williams\", \n\t\"statements\": [\n\t\t{\n\t\t\t\"statement\": \"NOT(sport-basketball) but (tennis) unlike 4 recorded winners of Best Female Athlete ESPY Award.\", \n\t\t\t\"score\": 0.36\n\t\t},\n  \t{\n\t\t\t\"statement\": \"NOT(occupation-politician) but (tennis player, businessperson, autobiographer) unlike 20 recorded  winners of Michigan Women's Hall of Fame.\",\n\t\t\t\"score\": 0.17\n\t\t}\n\t]\n}</code></pre>\n\n<p><strong>This data can be also browsed at:&nbsp;<a href=\"https://wikiknowledge.onrender.com/demographics/\">https://wikiknowledge.onrender.com/demographics/</a></strong></p>", 
    "language": "eng", 
    "title": "Wiki-based Communities of Interest: Demographics and Outliers", 
    "license": {
      "id": "CC-BY-4.0"
    }, 
    "relations": {
      "version": [
        {
          "count": 9, 
          "index": 8, 
          "parent": {
            "pid_type": "recid", 
            "pid_value": "7410436"
          }, 
          "is_last": true, 
          "last_child": {
            "pid_type": "recid", 
            "pid_value": "7537200"
          }
        }
      ]
    }, 
    "keywords": [
      "wikipedia", 
      "wikimedia", 
      "wikidata", 
      "demography", 
      "trivia"
    ], 
    "publication_date": "2022-12-07", 
    "creators": [
      {
        "affiliation": "Max Planck Institute for Informatics", 
        "name": "Hiba Arnaout"
      }, 
      {
        "affiliation": "Max Planck Institute for Informatics", 
        "name": "Simon Razniewski"
      }, 
      {
        "affiliation": "The University of Edinburgh", 
        "name": "Jeff Z. Pan"
      }
    ], 
    "access_right": "open", 
    "resource_type": {
      "type": "dataset", 
      "title": "Dataset"
    }, 
    "related_identifiers": [
      {
        "scheme": "doi", 
        "identifier": "10.5281/zenodo.7410436", 
        "relation": "isVersionOf"
      }
    ]
  }
}