Class: AIPerformanceMetrics¶

Performance metrics for AI assistance, extending general performance metrics with AI-specific measures

```mermaid classDiagram class AIPerformanceMetrics click AIPerformanceMetrics href "../AIPerformanceMetrics/" PerformanceMetrics <|-- AIPerformanceMetrics click PerformanceMetrics href "../PerformanceMetrics/"

  AIPerformanceMetrics : accuracy

  AIPerformanceMetrics : area_under_curve

  AIPerformanceMetrics : balanced_accuracy

  AIPerformanceMetrics : confidence_calibration

  AIPerformanceMetrics : diagnostic_odds_ratio

  AIPerformanceMetrics : f1_score

  AIPerformanceMetrics : false_discovery_rate

  AIPerformanceMetrics : false_negative_rate

  AIPerformanceMetrics : false_omission_rate

  AIPerformanceMetrics : false_positive_rate

  AIPerformanceMetrics : human_agreement_rate

  AIPerformanceMetrics : matthews_correlation

  AIPerformanceMetrics : negative_predictive_value

  AIPerformanceMetrics : precision

  AIPerformanceMetrics : processing_speed

  AIPerformanceMetrics : sensitivity

  AIPerformanceMetrics : specificity

```

Inheritance¶

PerformanceMetrics
- AIPerformanceMetrics

Slots¶

Name	Cardinality and Range	Description	Inheritance
processing_speed	0..1 Float	Average processing time per item (seconds)	direct
human_agreement_rate	0..1 Float	Rate of agreement with human reviewers	direct
confidence_calibration	0..1 Float	Calibration of AI confidence scores	direct
sensitivity	0..1 Float	True positive rate (recall)	PerformanceMetrics
specificity	0..1 Float	True negative rate	PerformanceMetrics
precision	0..1 Float	Positive predictive value	PerformanceMetrics
negative_predictive_value	0..1 Float	Negative predictive value	PerformanceMetrics
false_positive_rate	0..1 Float	Type I error rate	PerformanceMetrics
false_negative_rate	0..1 Float	Type II error rate	PerformanceMetrics
false_discovery_rate	0..1 Float	Expected proportion of false discoveries	PerformanceMetrics
false_omission_rate	0..1 Float	Proportion of false negatives among negative calls	PerformanceMetrics
accuracy	0..1 Float	Overall accuracy	PerformanceMetrics
balanced_accuracy	0..1 Float	Average of sensitivity and specificity	PerformanceMetrics
f1_score	0..1 Float	Harmonic mean of precision and recall	PerformanceMetrics
matthews_correlation	0..1 Float	Matthews correlation coefficient	PerformanceMetrics
diagnostic_odds_ratio	0..1 Float	Ratio of odds of positive test in diseased vs non-diseased	PerformanceMetrics
area_under_curve	0..1 Float	Area under the ROC curve	PerformanceMetrics

Usages¶

used by	used in	type	used
AIAssistance	ai_performance_metrics	range	AIPerformanceMetrics

Identifier and Mapping Information¶

Schema Source¶

from schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction

Mappings¶

Mapping Type	Mapped Value
self	revaise:AIPerformanceMetrics
native	revaise:AIPerformanceMetrics

LinkML Source¶

Direct¶

name: AIPerformanceMetrics
description: Performance metrics for AI assistance, extending general performance
  metrics with AI-specific measures
from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
is_a: PerformanceMetrics
slots:
- processing_speed
- human_agreement_rate
- confidence_calibration
slot_usage:
  processing_speed:
    name: processing_speed
    description: Average processing time per item (seconds)
    range: float
  human_agreement_rate:
    name: human_agreement_rate
    description: Rate of agreement with human reviewers
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  confidence_calibration:
    name: confidence_calibration
    description: Calibration of AI confidence scores
    range: float

Induced¶

name: AIPerformanceMetrics
description: Performance metrics for AI assistance, extending general performance
  metrics with AI-specific measures
from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
is_a: PerformanceMetrics
slot_usage:
  processing_speed:
    name: processing_speed
    description: Average processing time per item (seconds)
    range: float
  human_agreement_rate:
    name: human_agreement_rate
    description: Rate of agreement with human reviewers
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  confidence_calibration:
    name: confidence_calibration
    description: Calibration of AI confidence scores
    range: float
attributes:
  processing_speed:
    name: processing_speed
    description: Average processing time per item (seconds)
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - AIPerformanceMetrics
    range: float
  human_agreement_rate:
    name: human_agreement_rate
    description: Rate of agreement with human reviewers
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - AIPerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  confidence_calibration:
    name: confidence_calibration
    description: Calibration of AI confidence scores
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - AIPerformanceMetrics
    range: float
  sensitivity:
    name: sensitivity
    description: True positive rate (recall)
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  specificity:
    name: specificity
    description: True negative rate
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  precision:
    name: precision
    description: Positive predictive value
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  negative_predictive_value:
    name: negative_predictive_value
    description: Negative predictive value
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  false_positive_rate:
    name: false_positive_rate
    description: Type I error rate
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  false_negative_rate:
    name: false_negative_rate
    description: Type II error rate
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  false_discovery_rate:
    name: false_discovery_rate
    description: Expected proportion of false discoveries
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  false_omission_rate:
    name: false_omission_rate
    description: Proportion of false negatives among negative calls
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  accuracy:
    name: accuracy
    description: Overall accuracy
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  balanced_accuracy:
    name: balanced_accuracy
    description: Average of sensitivity and specificity
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  f1_score:
    name: f1_score
    description: Harmonic mean of precision and recall
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0
  matthews_correlation:
    name: matthews_correlation
    description: Matthews correlation coefficient
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: -1.0
    maximum_value: 1.0
  diagnostic_odds_ratio:
    name: diagnostic_odds_ratio
    description: Ratio of odds of positive test in diseased vs non-diseased
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
  area_under_curve:
    name: area_under_curve
    description: Area under the ROC curve
    from_schema: https://open-and-sustainable.github.io/revaise-model/schema/stages/extraction
    rank: 1000
    owner: AIPerformanceMetrics
    domain_of:
    - PerformanceMetrics
    range: float
    minimum_value: 0.0
    maximum_value: 1.0