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

71 lines
2.4 KiB
YAML

id: https://nde.nl/ontology/hc/class/ConfidenceMethod
name: confidence_method_class
title: Confidence Method
prefixes:
linkml: https://w3id.org/linkml/
hc: https://nde.nl/ontology/hc/
prov: http://www.w3.org/ns/prov#
schema: http://schema.org/
imports:
- linkml:types
- ../slots/has_description
- ../slots/identified_by
- ../slots/has_type
default_prefix: hc
classes:
ConfidenceMethod:
description: 'A method or algorithm used to calculate confidence scores.
**USAGE**: Documents how confidence values were computed: - Fuzzy string matching algorithms - ML model predictions
- Rule-based validation - XPath match verification - Human assessment
**COMMON METHODS**: | Method | Description | |--------|-------------| | fuzzy_matching | Levenshtein, Jaro-Winkler,
etc. | | xpath_validation | XPath match confidence | | llm_classification | LLM-based entity classification | | ml_prediction
| Machine learning model output | | human_assessment | Manual quality assessment | | ensemble | Combined multiple methods
|'
class_uri: prov:Plan
exact_mappings:
- prov:Plan
close_mappings:
- schema:HowTo
slots:
- has_type
- has_description
- identified_by
slot_usage:
has_type:
# range: string
required: true
examples:
- value: fuzzy_matching
- value: ml_prediction
- value: human_assessment
has_description:
# range: string
required: false
examples:
- value: Jaro-Winkler similarity with 0.7 threshold
identified_by:
# range: string
required: false
examples:
- value: rapidfuzz-2.15.1
annotations:
custodian_types: '["*"]'
custodian_types_rationale: Confidence methods apply universally to data quality assessment.
custodian_types_primary: '*'
specificity_score: 0.25
specificity_rationale: Low specificity - fundamental methodology documentation.
examples:
- value:
has_type: fuzzy_matching
has_description: Levenshtein distance with ratio normalization
identified_by: rapidfuzz-levenshtein
- value:
has_type: llm_classification
has_description: GPT-4 based entity type classification
identified_by: gpt-4-turbo-2024-04-09
comments:
- Created from slot_fixes.yaml migration (2026-01-19)
- Documents confidence calculation methodology
- Used with ConfidenceScore class