Skip to content

Commit

Permalink
adds Bias concepts to RISK and AI - see #182
Browse files Browse the repository at this point in the history
- adds bias taxnomy to RISK
- adds bias taxonomy (specific to AI) to AI
- contains RDF and HTML outputs

Co-authored-by: Daniel Doherty <[email protected]>
  • Loading branch information
coolharsh55 and drd00 committed Sep 18, 2024
1 parent 6acbf2b commit fa9eebd
Show file tree
Hide file tree
Showing 54 changed files with 68,191 additions and 32,392 deletions.
1,658 changes: 1,552 additions & 106 deletions 2.1-dev/ai/ai-en.html

Large diffs are not rendered by default.

1,530 changes: 1,463 additions & 67 deletions 2.1-dev/ai/ai-owl.html

Large diffs are not rendered by default.

1,443 changes: 1,154 additions & 289 deletions 2.1-dev/ai/ai-owl.jsonld

Large diffs are not rendered by default.

196 changes: 196 additions & 0 deletions 2.1-dev/ai/ai-owl.n3
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,7 @@
@prefix profile: <http://www.w3.org/ns/dx/prof/> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix risk-owl: <https://w3id.org/dpv/risk/owl#> .
@prefix role: <http://www.w3.org/ns/dx/prof/role/> .
@prefix schema: <https://schema.org/> .
@prefix skos: <http://www.w3.org/2004/02/skos/core#> .
Expand All @@ -25,6 +26,17 @@ ai-owl:AI a rdfs:Class,
skos:prefLabel "Artificial Intelligence (AI)"@en ;
skos:scopeNote "This concept is a stub"@en .

ai-owl:AIBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-18"^^xsd:date ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf risk-owl:Bias ;
sw:term_status "accepted"@en ;
skos:definition "Bias associated with development, use, or other activities involving an AI technology or system"@en ;
skos:prefLabel "AI Bias"@en .

ai-owl:AISystem a rdfs:Class,
owl:Class ;
dct:created "2024-05-17"^^xsd:date ;
Expand All @@ -40,6 +52,30 @@ ai-owl:AISystem a rdfs:Class,
skos:definition "OECD 2024 definition: An AI system is a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment. ISO/IEC 22989:2023 definition: engineered system that generates outputs such as content, forecasts, recommendations or decisions for a given set of human-defined objectives"@en ;
skos:prefLabel "AI System"@en .

ai-owl:AlgorithmSelectionBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from the selection of machine learning algorithms built into the AI system which introduce unwanted bias in predictions made by the system because the type of algorithm used introduces a variation in the performance of the ML model"@en ;
skos:prefLabel "Algorithm Selection Bias"@en .

ai-owl:AutomationBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf risk-owl:CognitiveBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias tha occurs due to propensity for humans to favour suggestions from automated decision-making systems and to ignore contradictory information made without automation, even if it is correct"@en ;
skos:prefLabel "Automation Bias"@en .

ai-owl:Capability a rdfs:Class,
owl:Class ;
dct:created "2024-04-28"^^xsd:date ;
Expand All @@ -50,6 +86,104 @@ ai-owl:Capability a rdfs:Class,
skos:prefLabel "Capability"@en ;
skos:scopeNote "This concept is a stub"@en .

ai-owl:DataAggregationBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:DataBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from aggregating data covering different groups of objects that might have different statistical distributions which introduce bias into the data used to train AI systems"@en ;
skos:prefLabel "Data Aggregation Bias"@en .

ai-owl:DataBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:AIBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs due to unaddressed data properties that lead to AI systems that perform better or worse for different groups"@en ;
skos:prefLabel "Data Bias"@en .

ai-owl:DataLabelsAndLabellingProcessBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:DataBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs due to the labelling process itself introducing societal or cognitive biases"@en ;
skos:prefLabel "Data Labels And Labelling Process Bias"@en .

ai-owl:DistributedTrainingBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:DataBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs due to distributed machine having different sources of data that do not have the same distribution of feature space"@en ;
skos:prefLabel "Distributed Training Bias"@en .

ai-owl:EngineeringDecisionBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:AIBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs due to machine learning model architectures - encompassing all model specifications, parameters and manually designed features"@en ;
skos:prefLabel "Engineering Decision Bias"@en ;
skos:scopeNote "Data bias and human cognitive bias can contribute to such bias"@en .

ai-owl:FeatureEngineeringBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from steps such as encoding, data type conversion, dimensionality reduction and feature selection which are subject to choices made by the AI developer and introduce bias in the ML model"@en ;
skos:prefLabel "Feature Engineering Bias"@en .

ai-owl:HyperparameterTuningBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from hyperparameters defining how the model is structured and which cannot be directly trained from the data like model parameters, where hyperparameters affect the model functioning and accuracy of the model"@en ;
skos:prefLabel "Hyperparameter Tuning Bias"@en .

ai-owl:InformativenessBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs or some groups, the mapping between inputs present in the data and outputs are more difficult to learn and where a model that only has one feature set available, can be biased against the group whose relationships are difficult to learn from available data"@en ;
skos:prefLabel "Informativeness Bias"@en ;
skos:scopeNote "This can happen when some features are highly informative about one group, while a different set of features is highly informative about another group. If this is the case, then a model that only has one feature set available, can be biased against the group whose relationships are difficult to learn from available data"@en .

ai-owl:Measure a rdfs:Class,
owl:Class ;
dct:created "2024-04-28"^^xsd:date ;
Expand All @@ -60,6 +194,19 @@ ai-owl:Measure a rdfs:Class,
skos:prefLabel "Measure"@en ;
skos:scopeNote "This concept is a stub"@en .

ai-owl:MissingFeaturesAndLabelsBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:DataBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs when features are missing from individual training samples"@en ;
skos:prefLabel "Missing Features And Labels Bias"@en ;
skos:scopeNote "If the frequency of missing features is higher for one group than another then this presents another vector for bias"@en .

ai-owl:Model a rdfs:Class,
owl:Class ;
dct:created "2024-05-17"^^xsd:date ;
Expand All @@ -70,6 +217,54 @@ ai-owl:Model a rdfs:Class,
skos:definition "Physical, mathematical or otherwise logical representation of a system, entity, phenomenon, process or data"@en ;
skos:prefLabel "Model"@en .

ai-owl:ModelBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs when ML uses functions like a maximum likelihood estimator to determine parameters, and there is data skew or under-representation present in the data, where the maximum likelihood estimation tends to amplify any underlying bias in the distribution"@en ;
skos:prefLabel "Model Bias"@en .

ai-owl:ModelExpressivenessBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:ModelInteractionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from the number and nature of parameters in a model as well as the neural network topology which affect the expressiveness of the model and any feature that affects model expressiveness differently across groups"@en ;
skos:prefLabel "Model Expressiveness Bias"@en .

ai-owl:ModelInteractionBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:EngineeringDecisionBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs from the structure of a model to create biased predictions"@en ;
skos:prefLabel "Model Interaction Bias"@en .

ai-owl:NonRepresentativeSamplingBias a rdfs:Class,
owl:Class,
risk-owl:RiskConcept ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-09-13"^^xsd:date ;
dct:source "ISO/IEC 24027:2021"@en ;
rdfs:isDefinedBy ai-owl: ;
rdfs:subClassOf ai-owl:DataBias ;
sw:term_status "accepted"@en ;
skos:definition "Bias that occurs if a dataset is not representative of the intended deployment environment, where the model learns biases based on the ways in which the data is non-representative"@en ;
skos:prefLabel "Non-Representative Sampling Bias"@en .

ai-owl:Risk a rdfs:Class,
owl:Class ;
dct:created "2024-04-28"^^xsd:date ;
Expand Down Expand Up @@ -152,6 +347,7 @@ ai-owl: a owl:Ontology,
dct:conformsTo <http://www.w3.org/2002/07/owl>,
"http://www.w3.org/2000/01/rdf-schema",
"http://www.w3.org/2004/02/skos/core" ;
dct:contributor "Daniel Doherty" ;
dct:created "2024-05-31"@en ;
dct:creator "Delaram Golpayegani"@en,
"Harshvardhan J. Pandit"@en ;
Expand Down
Loading

0 comments on commit fa9eebd

Please sign in to comment.