{
  "identifier": "tag:aim-pro.eu,2026:oer/ef26558dbe95/record",
  "profile": "aimpro-oer-profile/1",
  "normaliser": "oerlib.metadata/2",
  "status": "ready",
  "normalised_at": "2026-10-09T09:50:52+00:00",
  "resource": {
    "id": "ef26558dbe95",
    "identifier": "tag:aim-pro.eu,2026:oer/ef26558dbe95",
    "uri": "https://arxiv.org/abs/2310.18355",
    "title": "Health Disparities through Generative AI Models: A Comparison Study Using A Domain Specific large language model",
    "description": "Health disparities are differences in health outcomes and access to healthcare between different groups, including racial and ethnic minorities, low-income people, and rural residents. An artificial intelligence (AI) program called large language models (LLMs) can understand and generate human language, improving health communication and reducing health disparities. There are many challenges in using LLMs in human-doctor interaction, including the need for diverse and representative data, privacy concerns, and collaboration between healthcare providers and technology experts. We introduce the comparative investigation of domain-specific large language models such as SciBERT with a multi-purpose LLMs BERT. We used cosine similarity to analyze text queries about health disparities in exam rooms when factors such as race are used alone. Using text queries, SciBERT fails when it doesn't differentiate between queries text: \"race\" alone and \"perpetuates health disparities.\" We believe clinicians can use generative AI to create a draft response when communicating asynchronously with patients. However, careful attention must be paid to ensure they are developed and implemented ethically and equitably.",
    "creators": [
      "Yohn Jairo Parra Bautista",
      "Vinicious Lima",
      "Carlos Theran",
      "Richard Alo"
    ],
    "keywords": [
      "cs.CL",
      "cs.AI",
      "cs.LG"
    ],
    "kind": "html",
    "content_language": "en",
    "declared_language": "en",
    "word_count": 6065,
    "content_hash": "8052192407e25c61",
    "metadata_only": false
  },
  "licence": {
    "raw": "CC-BY-4.0",
    "spdx": "CC-BY-4.0",
    "uri": "http://creativecommons.org/licenses/by/4.0/",
    "verdict": "OPEN",
    "note": "open, permissive",
    "ingestable": true,
    "commercial_ok": true,
    "derivable": true,
    "share_alike": false,
    "evidence": {
      "scope": "this submitted version of the paper",
      "method": "arXiv OAI-PMH GetRecord (metadataPrefix=arXiv)",
      "schema": 1,
      "status": "captured",
      "location": "arXiv:arXiv/arXiv:license",
      "source_url": "https://oaipmh.arxiv.org/oai?verb=GetRecord&identifier=oai:arXiv.org:2310.18355&metadataPrefix=arXiv",
      "asserted_by": "arXiv submitter",
      "captured_at": "2026-10-09T09:50:49+00:00",
      "source_system": "arxiv",
      "observed_value": "http://creativecommons.org/licenses/by/4.0/",
      "documentation_url": "https://info.arxiv.org/help/oa/index.html",
      "responsibility_url": "https://info.arxiv.org/help/license/index.html",
      "documentation_label": "arXiv OAI-PMH metadata documentation",
      "responsibility_note": "The submitter chooses the article-version licence and certifies the right to grant it."
    }
  },
  "descriptive": {
    "contributions": [
      {
        "name": "Yohn Jairo Parra Bautista",
        "role": "author",
        "affiliation": "",
        "identifier": "",
        "scope": "resource"
      },
      {
        "name": "Vinicious Lima",
        "role": "author",
        "affiliation": "",
        "identifier": "",
        "scope": "resource"
      },
      {
        "name": "Carlos Theran",
        "role": "author",
        "affiliation": "",
        "identifier": "",
        "scope": "resource"
      },
      {
        "name": "Richard Alo",
        "role": "author",
        "affiliation": "",
        "identifier": "",
        "scope": "resource"
      }
    ],
    "identifiers": [
      {
        "scheme": "arxiv",
        "value": "2310.18355",
        "scope": "resource"
      }
    ],
    "dates": [
      {
        "meaning": "issued",
        "value": "2023-10-23",
        "scope": "resource"
      }
    ],
    "subjects": [
      {
        "value": "cs.CL",
        "scheme": "arxiv",
        "scope": "resource"
      },
      {
        "value": "cs.AI",
        "scheme": "arxiv",
        "scope": "resource"
      },
      {
        "value": "cs.LG",
        "scheme": "arxiv",
        "scope": "resource"
      }
    ],
    "relations": [],
    "educational": {
      "learning_resource_type": "",
      "intended_end_user_role": "",
      "context": "",
      "difficulty": "",
      "typical_learning_time": "",
      "description": ""
    },
    "container": {
      "kind": "",
      "uri": "",
      "title": "",
      "description": "",
      "identifier": ""
    },
    "declared_language": "en",
    "declared_type": "preprint",
    "rights_holder": "",
    "access_rights": "open",
    "publisher": "",
    "version": "v1",
    "citation": "",
    "notes": "",
    "lifecycle_status": "",
    "structure_role": "",
    "snapshot": {
      "resource": {
        "title": "Health Disparities through Generative AI Models: A Comparison Study Using A Domain Specific large language model",
        "authors": [
          "Yohn Jairo Parra Bautista",
          "Vinicious Lima",
          "Carlos Theran",
          "Richard Alo"
        ],
        "summary": "Health disparities are differences in health outcomes and access to healthcare between different groups, including racial and ethnic minorities, low-income people, and rural residents. An artificial intelligence (AI) program called large language models (LLMs) can understand and generate human language, improving health communication and reducing health disparities. There are many challenges in using LLMs in human-doctor interaction, including the need for diverse and representative data, privacy concerns, and collaboration between healthcare providers and technology experts. We introduce the comparative investigation of domain-specific large language models such as SciBERT with a multi-purpose LLMs BERT. We used cosine similarity to analyze text queries about health disparities in exam rooms when factors such as race are used alone. Using text queries, SciBERT fails when it doesn't differentiate between queries text: \"race\" alone and \"perpetuates health disparities.\" We believe clinicians can use generative AI to create a draft response when communicating asynchronously with patients. However, careful attention must be paid to ensure they are developed and implemented ethically and equitably.",
        "updated": "2023-10-23T21:24:05Z",
        "version": 1,
        "arxiv_id": "2310.18355",
        "published": "2023-10-23T21:24:05Z",
        "categories": [
          "cs.CL",
          "cs.AI",
          "cs.LG"
        ],
        "affiliations": [
          "",
          "",
          "",
          ""
        ],
        "primary_category": "cs.CL"
      }
    }
  },
  "representations": [
    {
      "role": "extracted",
      "media_type": "text/markdown",
      "size_bytes": 41194,
      "filename": "",
      "location": "",
      "source_format": "latexml-html",
      "converter": "arxiv-html"
    }
  ],
  "observed": {
    "language": {
      "method": "declared",
      "reason": "declared English",
      "accepted": true,
      "declared": "en",
      "detected": "en",
      "confidence": null,
      "review_required": false
    }
  },
  "collection": {
    "source": "arxiv",
    "source_label": "arXiv",
    "collected_at": "2026-10-09T09:50:52.655603+00:00",
    "query": "(all:\"artificial intelligence\" OR all:\"machine learning\" OR all:\"generative AI\" OR all:\"deep learning\" OR all:\"reinforcement learning\" OR all:\"large language model\") AND (all:\"AI concepts\" OR all:\"types of AI\" OR all:\"AI fundamentals\" OR all:\"recognizing AI\" OR all:\"recognising AI\" OR all:\"general versus narrow AI\" OR all:\"narrow AI\" OR all:\"general AI\" OR all:\"machine intelligence\" OR all:\"AI strengths and weaknesses\" OR all:\"traditional software\" OR all:\"rule-based systems\" OR all:\"introduction to AI\" OR all:\"introduction to artificial intelligence\" OR all:\"artificial intelligence introduction\" OR all:\"AI primer\" OR all:\"foundations of artificial intelligence\" OR all:\"overview of AI\" OR all:\"understanding AI\" OR all:\"history of AI\" OR all:\"AI essentials\" OR all:\"AI terminology\" OR all:\"metaphors for AI\" OR all:\"AI fundamental concepts\" OR all:\"AI key concepts\" OR all:\"philosophy of AI\" OR all:\"critical AI literacy\")",
    "note": "where this was collected from. Not a claim that the provider published or authored it"
  },
  "provenance": [
    {
      "field": "resource_id",
      "scope": "resource",
      "method": "engine",
      "evidence_ref": "engine#/resource_id"
    },
    {
      "field": "query",
      "scope": "resource",
      "method": "engine",
      "evidence_ref": "engine#/query"
    },
    {
      "field": "collected_from",
      "scope": "resource",
      "method": "engine",
      "evidence_ref": "engine#/source"
    },
    {
      "field": "title",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/title"
    },
    {
      "field": "description",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/summary"
    },
    {
      "field": "version",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource"
    },
    {
      "field": "subjects",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/categories/0"
    },
    {
      "field": "contributions",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/authors/0"
    },
    {
      "field": "contributions",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/authors/1"
    },
    {
      "field": "contributions",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/authors/2"
    },
    {
      "field": "contributions",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/authors/3"
    },
    {
      "field": "dates",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource"
    },
    {
      "field": "identifiers",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource/arxiv_id"
    },
    {
      "field": "content_language",
      "scope": "resource",
      "method": "gate_observation",
      "evidence_ref": "gate#/language/declared"
    },
    {
      "field": "declared_language",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource"
    },
    {
      "field": "licence",
      "scope": "resource",
      "method": "gate_observation",
      "evidence_ref": "gate#/licence/evidence"
    },
    {
      "field": "access_rights",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource"
    },
    {
      "field": "representations",
      "scope": "resource",
      "method": "conversion",
      "evidence_ref": "conversion#/latexml-html"
    },
    {
      "field": "declared_type",
      "scope": "resource",
      "method": "source_metadata",
      "evidence_ref": "snapshot#/resource"
    },
    {
      "field": "educational",
      "scope": "resource",
      "method": "declared_type_mapping",
      "evidence_ref": "profile#/learning_resource_type/preprint"
    }
  ],
  "incidents": [
    {
      "code": "absent",
      "field": "rights_holder",
      "detail": "no rights holder is named at source; the licence is recorded without one rather than attributed to the platform that served it"
    },
    {
      "code": "absent",
      "field": "publisher",
      "detail": "the source named no publisher of the work; where it was collected from is recorded as collection provenance instead, which is a different claim"
    },
    {
      "code": "absent",
      "field": "educational",
      "detail": "the source declared no educational metadata — no resource type, audience, context, difficulty or learning time. Nothing here estimates them"
    }
  ],
  "inferred": [
    {
      "rule": "purpose.declared_type",
      "field": "purpose",
      "value": "reference",
      "method": "inferred",
      "evidence": {
        "declared_type": "preprint"
      },
      "rule_version": "purpose/1"
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/2001.09346",
      "label": "arXiv:2001.09346",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 203,
            "basis": "reference",
            "literal": "arXiv:2001.09346"
          }
        ],
        "kind": "citation"
      }
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/2204.04521",
      "label": "arXiv:2204.04521",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 205,
            "basis": "reference",
            "literal": "arXiv:2204.04521"
          }
        ],
        "kind": "citation"
      }
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/2110.15621",
      "label": "arXiv:2110.15621",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 206,
            "basis": "reference",
            "literal": "arXiv:2110.15621"
          }
        ],
        "kind": "citation"
      }
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/1903.10676",
      "label": "arXiv:1903.10676",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 211,
            "basis": "reference",
            "literal": "arXiv:1903.10676"
          }
        ],
        "kind": "citation"
      }
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/2302.03494",
      "label": "arXiv:2302.03494",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 215,
            "basis": "reference",
            "literal": "arXiv:2302.03494"
          }
        ],
        "kind": "citation"
      }
    },
    {
      "field": "references",
      "value": "https://arxiv.org/abs/2304.03738",
      "label": "arXiv:2304.03738",
      "resource_id": "",
      "method": "inferred",
      "rule": "citation",
      "rule_version": "links/1",
      "evidence": {
        "links": [
          {
            "how": "citation",
            "line": 216,
            "basis": "reference",
            "literal": "arXiv:2304.03738"
          }
        ],
        "kind": "citation"
      }
    }
  ]
}