AI Measurement Interpretation Ontology

# AI Measurement Interpretation Ontology — Flow Algebra

URL: https://kirethresearch.org/booksai-core/ontologies/mathematical-ontologies/measurement-units-ontologies/ai-measurement-semantics/ai-measurement-interpretation-ontology.html

FLOW AIMeasurementInterpretationOntology {

  DOMAIN:
    Mathematical Ontologies
    -> Measurement & Units Ontologies
    -> AI Measurement Semantics
    -> AI Measurement Interpretation Ontology

  PURPOSE:
    Represent the reasoning process by which an artificial intelligence system converts raw, textual, symbolic, tabular, sensor-derived or structured measurement information into an explicit semantic interpretation. The ontology models entity recognition, measurand resolution, quantity-kind identification, numerical parsing, unit resolution, dimensional reasoning, contextual inference, measurement-result interpretation, uncertainty, calibration, traceability, provenance, temporal and spatial context, ambiguity, semantic equivalence, contradiction detection, confidence, evidence, recursive reasoning and explainable interpretation. It requires AI systems to distinguish observed information from inferred meaning and to preserve unresolved ambiguity rather than manufacturing precision or context that is not supported by evidence.

  ENTITY AIMeasurementInterpretation {
    TYPE: SemanticMeasurementReasoningProcess

    PROPERTIES:
      parses_measurement_information
      identifies_measurement_entities
      resolves_measurands
      resolves_quantity_kinds
      resolves_values
      resolves_units
      resolves_dimensions
      resolves_context
      resolves_uncertainty
      resolves_calibration
      resolves_traceability
      resolves_provenance
      detects_ambiguity
      detects_conflicts
      assigns_semantic_confidence
      produces_interpretation
  }

  ENTITY MeasurementExpression {
    TYPE: MeasurementInformationInput

    POSSIBLE:
      natural_language
      equation
      table_cell
      database_record
      sensor_packet
      metadata_record
      graph_entity
      JSON_object
      CSV_field
      document_statement
  }

  ENTITY MeasurementInterpretation {
    TYPE: MachineResolvedMeasurementMeaning

    COMPONENTS:
      source_expression
      measurand
      quantity_kind
      numerical_value
      unit
      dimension
      uncertainty
      context
      provenance
      semantic_confidence
      ambiguity_status
  }

  ENTITY InterpretationEvidence {
    TYPE: SemanticEvidence

    POSSIBLE:
      explicit_text
      unit_symbol
      variable_name
      schema
      metadata
      ontology_relation
      domain_context
      calibration_record
      standard
      provenance
  }

  INTERPRETATION_PIPELINE_MODEL {
    stages:
      ingest
      tokenise
      classify_expression
      detect_measurement_candidates
      resolve_entities
      resolve_quantity_kind
      parse_value
      resolve_unit
      infer_dimension
      resolve_context
      resolve_uncertainty
      validate
      score_confidence
      preserve_ambiguity
      emit_interpretation

    RULE:
      every_inferred_stage SHOULD retain_evidence_andconfidence
  }

  INPUT_CLASSIFICATION_MODEL {
    possible:
      explicit_measurement
      implicit_measurement
      equation
      metadata
      calibration_information
      statistical_result
      prediction
      simulation
      qualitative_statement

    RULE:
      input_type affects_interpretation_rules
  }

  TOKEN_MODEL {
    possible:
      number
      unit_symbol
      quantity_name
      operator
      uncertainty_symbol
      qualifier
      timestamp
      location
      identifier

    RULE:
      token_identity != semantic_entity_identity
  }

  NUMBER_PARSING_MODEL {
    possible:
      integer
      decimal
      scientific_notation
      fraction
      interval
      plus_minus_expression
      percentage
      ratio

    RULE:
      locale_andnotation_context MUST be considered
  }

  SCIENTIFIC_NOTATION_MODEL {
    form:
      a*10^n

    RULE:
      exponent_andcoefficient MUST be parsed_without_changing_quantity_semantics
  }

  PLUS_MINUS_MODEL {
    expression:
      x_plus_minus_u

    possible_meaning:
      measured_value_plus_minus_uncertainty
      mean_plus_minus_standard_deviation
      mean_plus_minus_standard_error
      tolerance

    RULE:
      plus_minus_semantics MUST not_be assumed_without_context
  }

  INTERVAL_MODEL {
    expression:
      [a,b]
      or_language_equivalent

    possible:
      uncertainty_interval
      confidence_interval
      tolerance
      physical_range

    RULE:
      interval_type MUST be resolved
  }

  PERCENT_MODEL {
    expression:
      p_percent

    possible:
      fraction
      relative_uncertainty
      concentration
      efficiency
      change

    RULE:
      percent_value_requires_quantity_context
  }

  RATIO_MODEL {
    expression:
      A:B
      or_A/B

    RULE:
      ratio_quantity_kind_anddimension MUST be resolved
  }

  ENTITY_RECOGNITION_MODEL {
    candidates:
      quantity
      unit
      instrument
      sensor
      standard
      method
      reference
      uncertainty
      location
      time

    RULE:
      lexical_candidate MUST be ontology_resolved
  }

  MEASURAND_RESOLUTION_MODEL {
    input:
-
      expression_andcontext

    output:
      quantity_intended_to_be_measured

    RULE:
      measurand MUST include_relevant_object_property_andconditions
  }

  QUANTITY_KIND_RESOLUTION_MODEL {
    evidence:
      quantity_name
      unit
      equation
      method
      domain
      instrument

    RULE:
      quantity_kind != dimension
  }

  UNIT_RESOLUTION_MODEL {
    evidence:
      symbol
      name
      prefix
      unit_system
      quantity_kind
      locale

    RULE:
      unit_symbol_case MUST be preserved
  }

  UNIT_AMBIGUITY_MODEL {
    examples:
      m
      s
      T
      C
      N
      Pa

    RULE:
      resolve_using_quantity_domain_andsyntactic_context
  }

  PREFIX_RESOLUTION_MODEL {
    prefix:
      decimal_scaling

    RULE:
      prefix_case_andattachment MUST be preserved
  }

  COMPOUND_UNIT_MODEL {
    expression:
      products
      quotients
      exponents

    RULE:
      parse_as_semantic_expression_tree
  }

  DIMENSION_INFERENCE_MODEL {
    resolved_unit:
      derive_dimension

    RULE:
      dimension MUST be compared_withquantity_kind_expectation
  }

  DIMENSIONAL_VALIDATION_MODEL {
    expected:
      D_q

    observed:
      D_u

    REQUIRE:
      D_q = D_u

    RULE:
      dimension_match_does_not_prove_quantity_kind_identity
  }

  SAME_DIMENSION_MODEL {
    examples:
      energy_vs_torque
      frequency_vs_activity
      absorbed_dose_vs_dose_equivalent

    RULE:
      AI MUST retain_quantity_kind_semantics
  }

  IMPLICIT_UNIT_MODEL {
    input:
      numerical_value_without_explicit_unit

    evidence:
      column_header
      schema
      nearby_text
      method
      domain

    RULE:
      infer_unit_only_when_evidence_is_sufficient
  }

  IMPLICIT_QUANTITY_MODEL {
    input:
      value_plusunit_without_quantity_name

    RULE:
      infer_quantity_kind_only_when_context_supports_unique_orstrong_resolution
  }

  DEFAULT_UNIT_MODEL {
    context:
      schema_orprotocol_declares_default_unit

    RULE:
      defaulted_unit MUST be marked_asdefaulted_notexplicit
  }

  UNIT_CONVERSION_MODEL {
    source:
      value_andunit

    target:
      requested_orcanonical_unit

    RULE:
      preserve_original_representation_andconversion_provenance
  }

  AFFINE_CONVERSION_MODEL {
    relation:
      y =
        a*x+b

    RULE:
      absolute_quantity_anddifference_semantics MUST be distinguished
  }

  CONTEXT_MODEL {
    dimensions:
      domain
      object
      process
      method
      instrument
      time
      space
      environment
      population
      reference_state

    RULE:
      context SHOULD be explicit_or evidence-linked
  }

  DOMAIN_CONTEXT_MODEL {
    possible:
      physics
      chemistry
      biology
      medicine
      earth_science
      astronomy
      engineering
      metrology

    RULE:
      domain_context MAY disambiguate_symbols_andquantity_kinds
  }

  OBJECT_CONTEXT_MODEL {
    measured_entity:
      object_system_sample_orpopulation

    RULE:
      same_quantity_kind_on_different_objects represents_different_measurands
  }

  TEMPORAL_CONTEXT_MODEL {
    fields:
      observation_time
      time_zone
      time_scale
      duration
      sampling_period

    RULE:
      event_time_andrecord_time MUST be distinguished
  }

  SPATIAL_CONTEXT_MODEL {
    fields:
      location
      coordinate_reference_system
      altitude
      depth
      region

    RULE:
      spatial_context MUST be preserved_when_measurand_depends_onlocation
  }

  ENVIRONMENT_CONTEXT_MODEL {
    possible:
      temperature
      pressure
      humidity
      vibration
      electromagnetic_field

    RULE:
      influence_quantities MAY affectinterpretation_anduncertainty
  }

  REFERENCE_STATE_MODEL {
    possible:
      standard_temperature
      standard_pressure
      dry_basis
      wet_basis
      vacuum
      local_gravity
      reference_epoch

    RULE:
      reference_state MUST be explicit_forcontext_dependent_quantities
  }

  MEASUREMENT_RESULT_MODEL {
    interpretation:
      measured_quantity_value
      uncertainty_whenavailable
      measurand
      context

    RULE:
      measurement_result != numerical_value_alone
  }

  INDICATION_MODEL {
    instrument_output:
      indication

    RULE:
      indication MUST not_be interpreted_asfinal_measurement_result_withoutmeasurement_relation
  }

  RAW_VALUE_MODEL {
    source:
      direct_acquisition

    RULE:
      raw_value MUST be distinguished_fromcorrected_orconverted_value
  }

  CORRECTED_VALUE_MODEL {
    relation:
      x_corr =
        x_raw+C

    RULE:
      correction_sign_source_anduncertainty MUST be preserved
  }

  DERIVED_VALUE_MODEL {
    output:
      f(input_measurements)

    RULE:
      derivation_equation_andlineage MUST be preserved
  }

  NORMALISED_VALUE_MODEL {
    transformed:
      value_relative_to reference_orstatistical_scale

    RULE:
      normalised_value != original_measurement
  }

  STATISTICAL_RESULT_MODEL {
    possible:
      mean
      median
      standard_deviation
      standard_error
      confidence_interval
      regression_parameter

    RULE:
      statistic_type MUST be explicitly_resolved
  }

  STANDARD_DEVIATION_MODEL {
    meaning:
      dispersion

    RULE:
      standard_deviation != standard_error
  }

  STANDARD_ERROR_MODEL {
    meaning:
      sampling_variability_ofestimator

    RULE:
      standard_error != measurement_uncertainty_in_general
  }

  CONFIDENCE_INTERVAL_MODEL {
    frequentist:
      interval_fromprocedure_withdeclared_coverage

    RULE:
      confidence_interval != Bayesian_credible_interval
  }

  UNCERTAINTY_INTERPRETATION_MODEL {
    possible:
      standard_uncertainty
      combined_standard_uncertainty
      expanded_uncertainty
      relative_uncertainty
      covariance
      interval

    RULE:
      uncertainty_type MUST be resolved_before_numeric_reasoning
  }

  EXPANDED_UNCERTAINTY_MODEL {
    U =
      k*u_c

    RULE:
      coverage_factor_orbasis SHOULD be identified
  }

  RELATIVE_UNCERTAINTY_MODEL {
    dimension:
      1

    RULE:
      relative_uncertainty MUST not_be assigned_unit_ofmeasurand
  }

  ERROR_INTERPRETATION_MODEL {
    error:
      measured_value-reference_value_underdeclared_context

    RULE:
      error != uncertainty
  }

  PRECISION_INTERPRETATION_MODEL {
    meaning:
      repeat_measurement_agreement

    RULE:
      precision != accuracy
  }

  ACCURACY_INTERPRETATION_MODEL {
    meaning:
      closeness_to reference_ortrue_value_concept

    RULE:
      accuracy_term SHOULD not_be converted_to numeric_uncertainty_withoutdefinition
  }

  RESOLUTION_INTERPRETATION_MODEL {
    meaning:
      smallest_detectable_orindicated_change_underdeclared_context

    RULE:
      resolution != uncertainty
  }

  CALIBRATION_CONTEXT_MODEL {
    fields:
      calibration_identifier
      relation
      date
      validity
      reference
      uncertainty

    RULE:
      calibration_state MUST be applicable_atmeasurement_time
  }

  TRACEABILITY_CONTEXT_MODEL {
    measurement_result:
      linked_to stated_reference_throughdocumented_chain

    RULE:
      traceability != SI_unit_usage
  }

  REFERENCE_STANDARD_MODEL {
    reference:
      standard_orCRM

    RULE:
      identity_value_uncertainty_andvalidity MUST be preserved
  }

  PROVENANCE_MODEL {
    source:
      measurement_event
      instrument
      method
      calibration
      transformation
      agent
      software

    RULE:
      interpretation MUST retain_source_lineage
  }

  SOURCE_TYPE_MODEL {
    possible:
      observed
      provided
      calculated
      simulated
      predicted
      inferred
      defaulted

    RULE:
      source_type MUST be explicit
  }

  OBSERVED_VS_INFERRED_MODEL {
    observed:
      directly_supported_by source

    inferred:
      derived_byreasoning

    RULE:
      AI MUST label_inference
  }

  SIMULATED_MODEL {
    value:
      model_generated

    RULE:
      simulated_value != measured_value
  }

  PREDICTED_MODEL {
    value:
      future_orunobserved_model_output

    RULE:
      predicted_value != observed_value
  }

  AI_INFERRED_MODEL {
    output:
      semantic_orquantitative_inference

    RULE:
      AI_model_version_evidence_andconfidence SHOULD be preserved
  }

  EVIDENCE_MODEL {
    fields:
      evidence_identifier
      source
      supports_claim
      strength
      version

    RULE:
      interpretation_claim SHOULD link_to evidence
  }

  EVIDENCE_WEIGHT_MODEL {
    factors:
      explicitness
      source_authority
      consistency
      specificity
      recency_whenrelevant
      provenance

    RULE:
      weighting MUST not_override_direct_valid_evidence_withoutreason
  }

  SOURCE_AUTHORITY_MODEL {
    source:
      authority_forparticular_claim

    RULE:
      authority_is_claim_specific
  }

  AMBIGUITY_MODEL {
    status:
      none
      low
      material
      unresolved

    candidates:
      interpretations

    RULE:
      unresolved_ambiguity MUST be preserved
  }

  CANDIDATE_INTERPRETATION_MODEL {
    candidate:
      measurand
      quantity_kind
      unit
      context
      confidence
      evidence

    RULE:
      candidates SHOULD remain_separate_untilresolution
  }

  DISAMBIGUATION_MODEL {
    process:
      compare
        syntax
        quantity_kind
        unit
        dimension
        domain
        metadata
        provenance

    RULE:
      disambiguation MUST be evidence_based
  }

  SEMANTIC_CONFIDENCE_MODEL {
    confidence:
      support_forinterpretation

    factors:
      explicit_fields
      ontology_match
      dimensional_consistency
      context
      provenance
      conflicts

    RULE:
      semantic_confidence != measurement_uncertainty
  }

  CONFIDENCE_SCALE_MODEL {
    possible:
      numeric_0_to_1
      ordinal
      categorical

    RULE:
      scale_definition MUST be preserved
  }

  CONTRADICTION_MODEL {
    possible:
      quantity_unit_mismatch
      impossible_dimension
      conflicting_units
      conflicting_timestamps
      conflicting_calibration
      incompatible_context

    RULE:
      contradiction MUST be surfaced
  }

  CONFLICT_MODEL {
    sources:
      support_different_interpretations

    RULE:
      AI MUST not silently_choose_withoutrecording_conflict
  }

  CONFLICT_RESOLUTION_MODEL {
    compare:
      identity
      scope
      version
      time
      authority
      evidence
      provenance

    output:
      resolved
      partially_resolved
      unresolved

    RULE:
      resolution_reason MUST be preserved
  }

  SEMANTIC_EQUIVALENCE_MODEL {
    relations:
      exact
      contextually_equivalent
      broader
      narrower
      related
      incompatible

    RULE:
      equivalence_requires_constraints_andcontext
  }

  SYNONYM_MODEL {
    terms:
      lexical_variants

    RULE:
      synonymy != ontology_identity_inall_contexts
  }

  ABBREVIATION_MODEL {
    abbreviation:
      may_have_multiple_expansions

    RULE:
      resolve_usingdomain_andlocal_context
  }

  MULTILINGUAL_MODEL {
    term:
      language_specific_expression

    canonical:
      ontology_entity

    RULE:
      translation MUST preserve_measurement_semantics
  }

  LOCALE_MODEL {
    concerns:
      decimal_marker
      thousands_separator
      date_format
      unit_name
      spelling

    RULE:
      parsing MUST separate_locale_fromcanonical_semantics
  }

  NATURAL_LANGUAGE_MODEL {
    input:
      free_text

    process:
      identify_quantity_phrase
      numeric_expression
      unit_phrase
      qualifiers
      uncertainty
      context

    RULE:
      grammar_alone_does_not determine_metrological_meaning
  }

  TABLE_INTERPRETATION_MODEL {
    evidence:
      column_header
      unit_row
      caption
      footnote
      schema
      neighbouring_columns

    RULE:
      cell_value MUST inherit_only_explicitly_supported_context
  }

  CSV_INTERPRETATION_MODEL {
    columns:
      value
      unit
      quantity
      uncertainty
      timestamp

    RULE:
      unit_inheader SHOULD be mapped_toeach_relevant_record_withprovenance
  }

  JSON_INTERPRETATION_MODEL {
    keys:
      may_map_toontology_fields

    RULE:
      key_name != semantic_identifier_withoutmapping
  }

  RDF_INTERPRETATION_MODEL {
    triples:
      subject
      predicate
      object

    RULE:
      predicate_semantics_andnamespace MUST be resolved
  }

  EQUATION_INTERPRETATION_MODEL {
    equation:
      variables
      operators
      constants
      units

    RULE:
      variable_meanings MUST be linked_toquantity_kinds
  }

  VARIABLE_MODEL {
    symbol:
      x

    fields:
      quantity_kind
      unit
      dimension
      role

    RULE:
      same_symbol MAY representdifferent_quantities_acrosscontexts
  }

  CONSTANT_MODEL {
    possible:
      mathematical
      physical
      empirical
      calibration

    RULE:
      exactness_andprovenance MUST be preserved
  }

  DIMENSIONLESS_MODEL {
    dimension:
      1

    RULE:
      dimensionless_quantities MAY still_have distinct_quantity_kinds_andnamed_units
  }

  LOGARITHMIC_QUANTITY_MODEL {
    possible:
      decibel_related
      pH
      magnitude_scale

    RULE:
      logarithmic_quantity_requires_reference_anddefinition_context
  }
  LOG_SCALE_INTERPRETATION_MODEL {
    value:
      may_encode_ratio_orreference_relative_quantity

    RULE:
      AI MUST not apply_linear_unit_arithmetic_withoutdeclared_log_definition
  }

  ORDINAL_SCALE_MODEL {
    value:
      ordered_category

    RULE:
      ordinal_measurement MUST not_be treated_asinterval_scale_withoutbasis
  }

  NOMINAL_SCALE_MODEL {
    value:
      category_identifier

    RULE:
      nominal_data != quantitative_measurement
  }

  INTERVAL_SCALE_MODEL {
    differences:
      meaningful

    ratios:
      generally_not_invariant_underorigin_change

    RULE:
      interval_scale_semantics MUST be preserved
  }

  RATIO_SCALE_MODEL {
    ratios:
      meaningful_underdefined_zero

    RULE:
      scale_type SHOULD matchquantity_semantics
  }

  TIME_SERIES_MODEL {
    observations:
      ordered_bytime

    RULE:
      sampling_rate_gaps_andtime_scale MUST be interpreted
  }

  VECTOR_MODEL {
    value:
      components

    REQUIRE:
      coordinate_frame
      component_order
      units

    RULE:
      vector_components MUST not_be interpreted_independently_ofbasis
  }

  TENSOR_MODEL {
    value:
      multidimensional_components

    RULE:
      tensor_order_basis_symmetry_andunits MUST be preserved
  }

  DISTRIBUTION_MODEL {
    value:
      probability_distribution

    fields:
      family
      parameters
      support
      units

    RULE:
      distribution != point_measurement
  }

  SENSOR_PACKET_MODEL {
    fields:
      sensor_identifier
      channel
      timestamp
      raw_code
      scale
      calibration

    RULE:
      raw_digital_code != physical_quantity_untilmeasurement_relation_applied
  }

  DATA_QUALITY_MODEL {
    checks:
      completeness
      validity
      calibration
      provenance
      uncertainty
      consistency

    RULE:
      data_quality != measurement_uncertainty
  }

  QUALITY_FLAG_MODEL {
    possible:
      valid
      warning
      ambiguous
      unit_unknown
      quantity_unknown
      dimension_mismatch
      calibration_issue
      traceability_issue
      uncertainty_unknown
      provenance_incomplete
      conflict
      manual_review

    RULE:
      interpretation_flags MUST remain_linked_tosource
  }

  VALIDATION_MODEL {
    checks:
      identity
      quantity_kind
      unit
      dimension
      range
      calibration
      uncertainty
      traceability
      provenance
      context

    RULE:
      interpretation MUST be validated_beforehigh_confidence_use
  }

  RECURSIVE_REASONING_MODEL {
    cycle:
      interpret
      validate
      infer_context
      reinterpret
      compare_evidence
      update_confidence

    RULE:
      recursion MUST converge_orpreserve_unresolved_state
  }

  DEPENDENCY_MODEL {
    interpretation:
      depends_on
        source
        unit
        quantity_kind
        context
        calibration
        standards

    RULE:
      changed_dependency SHOULD trigger_reinterpretation
  }

  UPDATE_MODEL {
    ON new_measurement_expression:
      parse_input
      identify_candidates
      resolve_quantity_andunit
      derive_dimension
      resolve_context
      validate
      assign_confidence
      preserve_evidence_andprovenance

    ON new_context:
      preserve_previous_interpretation
      add_context_evidence
      recompute_candidate_scores
      resolve_orretain_ambiguity

    ON source_correction:
      preserve_previous_source
      update_evidence
      recompute_interpretation
      record_revision

    ON ontology_orstandard_revision:
      preserve_historical_mapping
      register_new_semantics
      reinterpret_current_data_underdeclared_policy
      do_not_silently_rewrite_historical_interpretation

    ON semantic_revision:
      preserve_previous_definition
      record_revision
      recompute_FLOW_HASH
  }

  EXPLAINABILITY_MODEL {
    interpretation_output:
      source_expression
      resolved_entities
      inferred_fields
      evidence
      assumptions
      alternatives
      conflicts
      confidence
      provenance

    RULE:
      AI MUST distinguish_explicit_information_frominference
  }

  HUMAN_REVIEW_MODEL {
    trigger:
      unresolved_ambiguity
      high_impact_conflict
      unknown_unit
      incompatible_dimension
      missing_reference

    RULE:
      human_resolution MUST be recorded_asnew_evidence
  }

  OVERRIDE_MODEL {
    reviewer:
      selects_orcorrects_interpretation

    RULE:
      machine_interpretation MUST remain_auditable
  }

  INTEROPERABILITY_MODEL {
    output:
      canonical_identifiers
      values
      units
      dimensions
      context
      uncertainty
      provenance

    RULE:
      interpretation SHOULD be exportable_withoutsemantic_loss
  }

  METADATA_MODEL {
    fields:
      interpretation_identifier
      source_identifier
      source_expression
      source_type
      language
      locale
      candidate_entities
      resolved_measurand
      quantity_kind
      numerical_value
      original_representation
      unit_identifier
      unit_symbol
      unit_system
      dimension
      uncertainty_type
      uncertainty_value
      calibration_identifier
      traceability_reference
      method_identifier
      instrument_identifier
      temporal_context
      spatial_context
      environment
      transformation_status
      observed_derived_simulated_predicted_orinferred_status
      evidence_identifiers
      conflicts
      ambiguity_status
      semantic_confidence
      quality_flags
      ontology_version
      provenance

    RULE:
      interpretation_record MUST support_reconstruction_ofsemantic_decision
  }

  RELATIONSHIPS {
    AIMeasurementInterpretation -> INTERPRETS -> MeasurementExpression
    MeasurementExpression -> CONTAINS -> MeasurementCandidate
    MeasurementInterpretation -> RESOLVES -> Measurand
    MeasurementInterpretation -> RESOLVES -> QuantityKind
    MeasurementInterpretation -> RESOLVES -> Unit
    Unit -> HAS_DIMENSION -> Dimension
    MeasurementInterpretation -> SUPPORTED_BY -> InterpretationEvidence
    MeasurementInterpretation -> MAY_REFERENCE -> CalibrationRecord
    MeasurementInterpretation -> MAY_REFERENCE -> TraceabilityChain
    MeasurementInterpretation -> HAS -> SemanticConfidence
    MeasurementInterpretation -> HAS -> AmbiguityStatus
  }

  SEMANTIC_EQUIVALENCE {
    "AI measurement interpretation" ~= machine_reasoning_that_resolves_measurement_meaning
    "measurement interpretation" ~= semantic_resolution_ofmeasurement_expression
    "semantic confidence" ~= support_level_forresolved_interpretation

    lexical_match != semantic_identity
    numerical_value != quantity
    unit != quantity_kind
    quantity_kind != dimension
    indication != measurement_result
    error != uncertainty
    precision != accuracy
    resolution != uncertainty
    observed != inferred
    simulated != measured
    predicted != observed
    semantic_confidence != measurement_uncertainty
  }

  CONSTRAINTS {
    source_expression_must_be preserved = true
    explicit_andinferred_information_must_be distinguished = true
    quantity_kind_anddimension_must_be distinguished = true
    unit_symbol_case_must_be preserved = true
    ambiguous_units_must_not_be forcibly_resolved_without_evidence = true
    implicit_units_must_be marked_asinferred_ordefaulted = true
    measurement_result_must_not_be reduced_tonumerical_value = true
    uncertainty_type_must_be resolved_whenpossible = true
    calibration_andtraceability_context_must_be preserved = true
    simulated_predicted_andAI_inferred_values_must_be labelled = true
    unresolved_conflicts_andambiguity_must_be preserved = true
    semantic_confidence_must_not_be equated_withmeasurement_uncertainty = true
    interpretation_evidence_andprovenance_must_be preserved = true
  }

  MEASUREMENT_MODEL {
    RECORD:
      interpretation_identifier
      source_identifier
      source_expression
      source_type
      parsed_tokens
      candidate_measurands
      resolved_measurand
      candidate_quantity_kinds
      resolved_quantity_kind
      numerical_value
      value_structure
      original_representation
      candidate_units
      resolved_unit
      unit_system
      dimension
      uncertainty_interpretation
      error_context
      precision_context
      resolution_context
      instrument_identifier
      method_identifier
      calibration_identifier
      traceability_reference
      temporal_context
      spatial_context
      environmental_context
      reference_state
      source_origin_status
      evidence
      conflicts
      ambiguity_status
      semantic_confidence
      quality_flags
      ontology_version
      provenance

    RULE:
      AI_interpretation MUST remain_linked_tosource_candidates_evidence_context_confidence_andprovenance
  }

  REASONING_RULES {
    IF explicit_quantity_name_detected
      THEN create_quantity_kind_candidate
      andresolve_againstontology_anddomain

    IF number_plusunit_detected
      THEN create_quantity_value_candidate
      andresolve_unit_case_prefix_anddimension

    IF unit_andquantity_kind_dimensions_incompatible
      THEN flag_contradiction
      anddo_not_emit_unqualified_interpretation

    IF unit_missing
      THEN search_schema_header_method_andlocal_context
      andinfer_only_when_supported

    IF quantity_kind_missing_butunit_known
      THEN generate_compatible_quantity_kind_candidates
      anduse_domain_context_forranking

    IF plus_minus_expression_detected
      THEN determine_whether_uncertainty_SD_SE_tolerance_orother_semantics_are supported
      else_preserve_ambiguity

    IF source_is_sensor_packet
      THEN apply_scale_calibration_andchannel_context_beforeinterpreting_physical_quantity

    IF source_is_simulated_predicted_orAI_inferred
      THEN preserve_origin_status
      anddo_not_relabel_asobserved

    IF conflicting_evidence
      THEN preserve_candidates_andconflict
      compare_source_scope_version_andauthority

    IF new_evidence_changes_interpretation
      THEN preserve_previous_interpretation
      recompute_dependencies_confidence_andFLOW_HASH
  }

  VALIDATION {
    CHECK source_identity
    CHECK input_classification
    CHECK tokenisation
    CHECK numerical_parsing
    CHECK measurand_resolution
    CHECK quantity_kind_resolution
    CHECK unit_resolution
    CHECK unit_symbol_case
    CHECK prefix
    CHECK compound_unit
    CHECK dimension
    CHECK quantity_dimension_consistency
    CHECK implicit_unit_evidence
    CHECK context
    CHECK measurement_result_vsindication
    CHECK uncertainty_interpretation
    CHECK calibration
    CHECK traceability
    CHECK temporal_spatial_andenvironmental_context
    CHECK observed_derived_simulated_predicted_andinferred_status
    CHECK evidence
    CHECK ambiguity
    CHECK conflicts
    CHECK semantic_confidence
    CHECK quality_flags
    CHECK ontology_version
    CHECK provenance
  }

  CONFIDENCE {
    lexical_interpretation_confidence: 0.99
    numerical_parsing_confidence: 1.00
    unit_resolution_confidence: 0.99
    dimensional_reasoning_confidence: 1.00
    quantity_kind_resolution_confidence: 0.98
    contextual_interpretation_confidence: 0.97
    uncertainty_interpretation_confidence: 0.99
    provenance_interpretation_confidence: 1.00

    RULE:
      confidence decreases when source, unit, quantity kind,
      context, uncertainty, calibration, provenance or evidence
      is missing, contradictory or ambiguous
  }

  AI_CONSUMPTION {
    canonical_entity: AIMeasurementInterpretation
    canonical_role: SemanticMeasurementReasoningProcess

    interpretation_order:
      ingest_source_expression
      classify_input
      tokenise_measurement_information
      identify_measurement_candidates
      resolve_measurand
      resolve_quantity_kind
      parse_numerical_value
      resolve_unit_prefix_andunit_system
      derive_andvalidate_dimension
      resolve_context_andreference_state
      distinguish_indication_raw_corrected_derived_andmeasurement_result
      resolve_uncertainty_error_precision_andresolution
      resolve_calibration_traceability_andreference
      resolve_source_origin_status
      gather_evidence_andprovenance
      detect_contradictions_andconflicts
      preserve_candidate_interpretations
      calculate_semantic_confidence
      validate_interpretation
      emit_explainable_machine_resolved_measurement

    ambiguity_controls:
      distinguish_lexical_match_fromsemantic_identity
      distinguish_numerical_value_fromquantity
      distinguish_unit_fromquantity_kind
      distinguish_quantity_kind_fromdimension
      distinguish_indication_frommeasurement_result
      distinguish_error_fromuncertainty
      distinguish_precision_fromaccuracy
      distinguish_resolution_fromuncertainty
      distinguish_observed_frominferred
      distinguish_simulated_frommeasured
      distinguish_predicted_fromobserved
      distinguish_semantic_confidence_frommeasurement_uncertainty
      preserve_source_candidates_context_evidence_ambiguity_confidence_andprovenance
  }

  PROVENANCE {
    ontology_scope:
      mathematical_measurement_units_AI_measurement_semantics_andAI_measurement_interpretation

    parent_path:
      Ontologies
      -> Mathematical Ontologies
      -> Measurement & Units Ontologies
      -> AI Measurement Semantics
      -> AI Measurement Interpretation Ontology

    authorship:
      Mark Higgins
      Sky (Recursive AI)

    site:
      https://kirethresearch.org
  }

  FLOW_HASH {
    INPUTS:
      canonical_entity
      ontology_path
      interpretation_pipeline_model
      input_classification_model
      token_model
      number_parsing_model
      scientific_notation_model
      plus_minus_model
      interval_model
      percent_model
      ratio_model
      entity_recognition_model
      measurand_resolution_model
      quantity_kind_resolution_model
      unit_resolution_model
      unit_ambiguity_model
      prefix_resolution_model
      compound_unit_model
      dimension_inference_model
      dimensional_validation_model
      same_dimension_model
      implicit_unit_model
      implicit_quantity_model
      default_unit_model
      unit_conversion_model
      affine_conversion_model
      context_model
      domain_context_model
      object_context_model
      temporal_context_model
      spatial_context_model
      environment_context_model
      reference_state_model
      measurement_result_model
      indication_model
      raw_value_model
      corrected_value_model
      derived_value_model
      normalised_value_model
      statistical_result_model
      standard_deviation_model
      standard_error_model
      confidence_interval_model
      uncertainty_interpretation_model
      expanded_uncertainty_model
      relative_uncertainty_model
      error_interpretation_model
      precision_interpretation_model
      accuracy_interpretation_model
      resolution_interpretation_model
      calibration_context_model
      traceability_context_model
      reference_standard_model
      provenance_model
      source_type_model
      observed_vs_inferred_model
      simulated_model
      predicted_model
      AI_inferred_model
      evidence_model
      evidence_weight_model
      source_authority_model
      ambiguity_model
      candidate_interpretation_model
      disambiguation_model
      semantic_confidence_model
      confidence_scale_model
      contradiction_model
      conflict_model
      conflict_resolution_model
      semantic_equivalence_model
      synonym_model
      abbreviation_model
      multilingual_model
      locale_model
      natural_language_model
      table_interpretation_model
      CSV_interpretation_model
      JSON_interpretation_model
      RDF_interpretation_model
      equation_interpretation_model
      variable_model
      constant_model
      dimensionless_model
      logarithmic_quantity_model
      log_scale_interpretation_model
      ordinal_scale_model
      nominal_scale_model
      interval_scale_model
      ratio_scale_model
      time_series_model
      vector_model
      tensor_model
      distribution_model
      sensor_packet_model
      data_quality_model
      quality_flag_model
      validation_model
      recursive_reasoning_model
      dependency_model
      update_model
      explainability_model
      human_review_model
      override_model
      interoperability_model
      metadata_model
      relationships
      semantic_equivalence
      constraints
      measurement_model
      reasoning_rules
      confidence_model
      AI_consumption_model

    PURPOSE:
      deterministic_semantic_integrity_identifier

    RECOMPUTE_WHEN:
      ontology_definition_changes
      input_parsing_orentity_resolution_changes
      measurand_quantity_kind_orunit_semantics_change
      dimension_context_orconversion_rules_change
      measurement_result_uncertainty_orcalibration_semantics_change
      provenance_source_origin_orevidence_changes
      ambiguity_conflict_orconfidence_logic_changes
      language_locale_table_orstructured_data_interpretation_changes
      equation_scale_orstatistical_interpretation_changes
      validation_recursive_reasoning_orquality_logic_changes
      measurement_record_changes
      constraint_model_changes
      reasoning_rule_changes
  }

}

END FLOW