glam/schemas/20251121/linkml/modules/classes/PersonWebClaim.yaml

id: https://nde.nl/ontology/hc/class/PersonWebClaim
name: PersonWebClaim
title: Person Web Claim - Verifiable Person Data Claims
prefixes:
  linkml: https://w3id.org/linkml/
  hc: https://nde.nl/ontology/hc/
  schema: http://schema.org/
  dcterms: http://purl.org/dc/terms/
  prov: http://www.w3.org/ns/prov#
  pico: https://personsincontext.org/model#
  foaf: http://xmlns.com/foaf/0.1/
imports:
- linkml:types
- ../slots/source_url
- ../slots/retrieved_on
- ../slots/retrieval_agent
- ../slots/person_claim_id
- ../slots/person_claim_note
- ../slots/person_claim_type
- ../slots/person_claim_value
- ../slots/person_html_file
- ../slots/person_xpath
- ../slots/person_xpath_match_score
- ../slots/specificity_annotation
- ../slots/template_specificity
- ./SpecificityAnnotation
- ./TemplateSpecificityScores
- ../enums/PersonClaimTypeEnum
- ../enums/RetrievalAgentEnum
default_prefix: hc
classes:
  PersonWebClaim:
    class_uri: prov:Entity
    description: "A verifiable claim about a person extracted from a web page with provenance.\n\n**RULE 26 COMPLIANCE: Person\
      \ Data Provenance**\n\nAll person/staff data associated with heritage custodians MUST have\nweb claim provenance. This\
      \ includes:\n- Staff names, titles, and roles\n- Contact information (if publicly available)\n- Professional history\
      \ and education\n- Affiliations and expertise areas\n\n**VERIFIABILITY PRINCIPLE**\n\nLike the base WebClaim class,\
      \ PersonWebClaim follows the \"XPath or Remove\" principle:\n- Claims from web pages MUST have xpath pointing to source\
      \ element\n- Claims from APIs (LinkedIn via Exa) may have xpath=null but MUST have source_url\n- Claims without any\
      \ verifiable source are FABRICATED and must be removed\n\n**SOURCE HIERARCHY**\n\nWhen multiple sources provide the\
      \ same information:\n1. Official institutional website (highest reliability)\n2. LinkedIn profile (high reliability)\n\
      3. News articles/press releases (medium-high)\n4. Conference programs (medium)\n5. Academic publications (medium)\n\
      6. Third-party databases (lower)\n\nDocument all sources when available; note conflicts.\n\n**CLAIM TYPES**\n\nCommon\
      \ claim types for heritage institution staff:\n- **full_name**: Complete name as displayed\n- **role_title**: Job title\
      \ (e.g., \"Senior Curator\")\n- **department**: Organizational unit\n- **email**: Professional contact (only if public)\n\
      - **biography**: Professional bio text\n- **specialization**: Expertise areas\n- **education**: Degrees and institutions\n\
      \n**EXTRACTION WORKFLOW**\n\n1. SCRAPE institutional staff/team pages with FireCrawl\n2. EXTRACT names and roles with\
      \ XPath locations\n3. SEARCH LinkedIn for additional profile data\n4. CREATE PersonWebClaim for each extracted fact\n\
      5. LINK claims to PersonObservation records\n\n**INTEGRATION WITH PERSONOBSERVATION**\n\nPersonObservation (the PiCo-based\
      \ staff role record) references\nPersonWebClaim instances via the `web_claims` slot:\n\n```yaml\nPersonObservation:\n\
      \  person_name: \"Dr. Jane Smith\"\n  staff_role: CONSERVATOR\n  has_or_had_web_claim:\n    - claim_type: full_name\n\
      \      claim_value: \"Dr. Jane Smith\"\n      source_url: https://museum.org/team\n      xpath: /html/body/main/div[2]/h3\n\
      \      retrieval_agent: firecrawl\n```\n\n**LINKEDIN PROFILE HANDLING**\n\nFor LinkedIn data, create separate profile\
      \ files (per Rule 12, Rule 20)\nand reference them:\n\n```yaml\nlinkedin_claims:\n  linkedin_url: https://www.linkedin.com/in/jane-smith\n\
      \  profile_data_path: data/custodian/person/entity/jane-smith_20250115.json\n  retrieved_on: \"2025-01-15T10:30:00Z\"\
      \n  retrieval_agent: exa_crawling_exa\n```\n"
    exact_mappings:
    - prov:Entity
    close_mappings:
    - schema:PropertyValue
    - foaf:Document
    slots:
    - person_claim_id
    - person_claim_note
    - person_claim_type
    - person_claim_value
    - person_html_file
    - person_xpath
    - person_xpath_match_score
    - retrieval_agent
    - retrieved_on
    - source_url
    - specificity_annotation
    - template_specificity
    slot_usage:
      person_claim_type:
        required: true
        description: 'Type of person claim. See PersonClaimTypeEnum.


          Common claim types:

          - full_name: Complete name

          - role_title: Job title

          - department: Organizational unit

          - email: Contact email (if public)

          - biography: Professional bio

          - specialization: Expertise areas

          '
      person_claim_value:
        required: true
        description: 'The extracted value. Must be verifiable at the source.

          '
      source_url:
        required: true
        description: 'URL where this claim was found.

          Required for ALL claims - enables verification.

          '
      retrieved_on:
        required: true
        description: 'Timestamp when data was retrieved.

          ISO 8601 format with timezone (UTC preferred).

          '
      retrieval_agent:
        required: true
        range: RetrievalAgentEnum
        description: 'Tool used to extract this data.


          - firecrawl: Institutional websites (primary)

          - playwright: JS-heavy sites

          - exa_crawling_exa: LinkedIn profiles

          - manual: Last resort

          '
      specificity_annotation:
        range: SpecificityAnnotation
        inlined: true
      template_specificity:
        range: TemplateSpecificityScores
        inlined: true
    comments:
    - PersonWebClaim extends WebClaim pattern for person-specific data
    - 'XPATH OR REMOVE: Claims without verifiable source must be removed'
    - Links to PersonObservation via web_claims slot
    - LinkedIn data stored separately in person/entity/ files (Rule 12, 20)
    - See Rule 26 in AGENTS.md for complete documentation
    examples:
    - value:
        person_claim_type: full_name
        person_claim_value: Taco Dibbits
        source_url: https://www.rijksmuseum.nl/en/about-us/organisation
        retrieved_on: '2025-01-15T10:30:00Z'
        person_xpath: /html/body/main/section[2]/div[1]/h2
        person_html_file: web/NL-NH-AMS-M-RM/rijksmuseum.nl/organisation.html
        person_xpath_match_score: 1.0
        retrieval_agent: firecrawl
      description: Exact match for museum director name
    - value:
        person_claim_type: role_title
        person_claim_value: General Director
        source_url: https://www.rijksmuseum.nl/en/about-us/organisation
        retrieved_on: '2025-01-15T10:30:00Z'
        person_xpath: /html/body/main/section[2]/div[1]/p[1]
        person_html_file: web/NL-NH-AMS-M-RM/rijksmuseum.nl/organisation.html
        person_xpath_match_score: 1.0
        retrieval_agent: firecrawl
      description: Role title from institutional page
    - value:
        person_claim_type: biography
        person_claim_value: Taco Dibbits has been General Director since 2016...
        source_url: https://www.rijksmuseum.nl/en/about-us/organisation
        retrieved_on: '2025-01-15T10:30:00Z'
        person_xpath: /html/body/main/section[2]/div[1]/div[2]
        person_html_file: web/NL-NH-AMS-M-RM/rijksmuseum.nl/organisation.html
        person_xpath_match_score: 0.92
        retrieval_agent: firecrawl
        person_claim_notes: Biography truncated from longer text on page
      description: Biography text with partial match score
    - value:
        person_claim_type: linkedin_url
        person_claim_value: https://www.linkedin.com/in/taco-dibbits
        source_url: https://www.linkedin.com/in/taco-dibbits
        retrieved_on: '2025-01-15T11:00:00Z'
        person_xpath: null
        retrieval_agent: exa_crawling_exa
        person_claim_notes: Profile data stored in person/entity/taco-dibbits_20250115.json
      description: LinkedIn claim - XPath null for API extraction