{
  "@context": {
    "@language": "en",
    "@vocab": "https://schema.org/",
    "citeAs": "cr:citeAs",
    "column": "cr:column",
    "conformsTo": "dct:conformsTo",
    "cr": "http://mlcommons.org/croissant/",
    "rai": "http://mlcommons.org/croissant/RAI/",
    "data": {
      "@id": "cr:data",
      "@type": "@json"
    },
    "dataType": {
      "@id": "cr:dataType",
      "@type": "@vocab"
    },
    "dct": "http://purl.org/dc/terms/",
    "examples": {
      "@id": "cr:examples",
      "@type": "@json"
    },
    "extract": "cr:extract",
    "field": "cr:field",
    "fileProperty": "cr:fileProperty",
    "fileObject": "cr:fileObject",
    "fileSet": "cr:fileSet",
    "format": "cr:format",
    "includes": "cr:includes",
    "isLiveDataset": "cr:isLiveDataset",
    "jsonPath": "cr:jsonPath",
    "key": "cr:key",
    "md5": "cr:md5",
    "parentField": "cr:parentField",
    "path": "cr:path",
    "recordSet": "cr:recordSet",
    "references": "cr:references",
    "regex": "cr:regex",
    "repeated": "cr:repeated",
    "replace": "cr:replace",
    "sc": "https://schema.org/",
    "separator": "cr:separator",
    "source": "cr:source",
    "subField": "cr:subField",
    "transform": "cr:transform"
  },
  "@type": "sc:Dataset",
  "name": "naijahealth-benchmark",
  "description": "A healthcare benchmarking dataset for evaluating LLM performance on medical queries across three Nigerian languages — Pidgin, Hausa, and Igbo. Each entry pairs an English health prompt (single-turn or multi-turn clinical conversations) with human translations, alongside a scored rubric tailored to the Nigerian healthcare context covering local disease prevalence, available resources, and clinical practice norms.",
  "conformsTo": "http://mlcommons.org/croissant/1.0",
  "license": "https://creativecommons.org/licenses/by-nc-nd/4.0/",
  "version": "0.1.0-preview",
  "comment": "This is a preview sample of a larger commercial dataset. For access to the full dataset, please contact the authors.",
  "inLanguage": ["en", "pcm", "ha", "ig"],
  "keywords": [
    "healthcare",
    "multilingual",
    "Nigeria",
    "Pidgin",
    "Hausa",
    "Igbo",
    "LLM evaluation",
    "benchmarking",
    "medical NLP",
    "low-resource languages"
  ],
  "distribution": [
    {
      "@type": "cr:FileObject",
      "@id": "data-file",
      "name": "naijahealth_output.json",
      "contentUrl": "naijahealthbench.json",
      "encodingFormat": "application/json"
    }
  ],
  "recordSet": [
    {
      "@type": "cr:RecordSet",
      "@id": "prompts",
      "name": "prompts",
      "description": "Healthcare prompts with multilingual translations and Nigerian-context scoring rubrics.",
      "field": [
        {
          "@type": "cr:Field",
          "@id": "prompts/prompt_id",
          "name": "prompt_id",
          "description": "Unique identifier (UUID) for the prompt scenario.",
          "dataType": "sc:Text",
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].prompt_id" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "prompts/prompt_text",
          "name": "prompt_text",
          "description": "English prompt text(s). Single-turn prompts have one entry; multi-turn conversations have multiple entries with labels (Prompt_1, Prompt_2, etc.).",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].prompts[*].text" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "prompts/prompt_label",
          "name": "prompt_label",
          "description": "Label indicating the prompt turn (e.g. 'Prompt', 'Prompt_1', 'Prompt_2').",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].prompts[*].label" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "prompts/pidgin",
          "name": "pidgin_translation",
          "description": "Nigerian Pidgin translation of the prompt(s).",
          "dataType": "sc:Text",
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].languages.Pidgin" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "prompts/hausa",
          "name": "hausa_translation",
          "description": "Hausa translation of the prompt(s).",
          "dataType": "sc:Text",
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].languages.Hausa" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "prompts/igbo",
          "name": "igbo_translation",
          "description": "Igbo translation of the prompt(s).",
          "dataType": "sc:Text",
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].languages.Igbo" }
          }
        }
      ]
    },
    {
      "@type": "cr:RecordSet",
      "@id": "rubric",
      "name": "rubric",
      "description": "Nigerian-context scoring rubric criteria with positive and negative scores.",
      "field": [
        {
          "@type": "cr:Field",
          "@id": "rubric/prompt_id",
          "name": "prompt_id",
          "description": "The prompt this rubric criterion belongs to.",
          "dataType": "sc:Text",
          "references": { "field": { "@id": "prompts/prompt_id" } },
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].prompt_id" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rubric/criteria",
          "name": "criteria",
          "description": "Rubric criterion text, adapted to the Nigerian healthcare context.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].rubric[*].criteria" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "rubric/score",
          "name": "score",
          "description": "Score weight for the criterion. Positive values reward presence; negative values penalize absence or errors.",
          "dataType": "sc:Float",
          "repeated": true,
          "source": {
            "fileObject": { "@id": "data-file" },
            "extract": { "jsonPath": "$[*].rubric[*].score" }
          }
        }
      ]
    }
  ]
}