language en

The Explainer Ontology

Latest version:
https://www.w3id.org/iSeeOnto/explainer
Contributors:
Anjana Wijekoon
Chamath Palihawadana
David Corsar
Ikechukwu Nkisi-Orji
Juan A. Recio-Garcia
Marta Caro Martínez
Imported Ontologies:
explanationPattern.owl
sio.owl
cpannotationschema.owl
prov-o#
eo
Download serialization:
RDF/XML
Visualization:
Visualize with WebVowl
Cite as:
The Explainer Ontology.
Ontology Specification Draft

Abstract

The Explainer Ontology is an ontology that models the technique used by an Explainer (as defined by the Explanation Ontology), to generate an Explanation. The concepts in this ontology reflect the a subset of the explanability fact sheet dimensions defined by Kacper Sokol and Peter Flach. 2020. Explainability fact sheets: a framework for systematic assessment of explainable approaches, in Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency (FAT* '20) DOI:https://doi.org/10.1145/3351095.3372870. This ontology was created as part of the iSee project (https://isee4xai.com) which received funding from EPSRC under the grant number EP/V061755/1. iSee is part of the CHIST-ERA pathfinder programme for European coordinated research on future and emerging information and communication technologies.

Introduction back to ToC

Namespace declarations

Table 1: Namespaces used in the document
explnr<https://www.w3id.org/iSeeOnto/explainer>
schema<http://schema.org>
explanationPattern<http://linkedu.eu/dedalo/explanationPattern.owl>
void<http://rdfs.org/ns/void>
owl<http://www.w3.org/2002/07/owl>
resource<http://semanticscience.org/resource>
xsd<http://www.w3.org/2001/XMLSchema>
protege<http://protege.stanford.edu/plugins/owl/protege>
cpannotationschema<http://www.ontologydesignpatterns.org/schemas/cpannotationschema.owl>
skos<http://www.w3.org/2004/02/skos/core>
eo<https://purl.org/heals/eo>
rdfs<http://www.w3.org/2000/01/rdf-schema>
cito<http://purl.org/spar/cito>
prov-o<http://www.w3.org/TR/prov-o>
rdf<http://www.w3.org/1999/02/22-rdf-syntax-ns>
terms<http://purl.org/dc/terms>
xml<http://www.w3.org/XML/1998/namespace>
explnr<http://www.isee4xai.com/ontologies/iseeonto/explainer>
vann<http://purl.org/vocab/vann>
obo<http://purl.obolibrary.org/obo>
prov<http://www.w3.org/ns/prov>
foaf<http://xmlns.com/foaf/0.1>
explainer<http://www.w3id.org/iSeeOnto/explainer>
dc<http://purl.org/dc/elements/1.1>

The Explainer Ontology: Overview back to ToC

This ontology has the following classes and properties.

Classes

Object Properties

Named Individuals

The Explainer Ontology: Description back to ToC

Outline of the Explainer ontology main classes and relationships
Outline of the Explainer ontology main classes and relationships. Concepts highlighted in peach are defined in this ontology.

Cross reference for The Explainer Ontology classes, properties and dataproperties back to ToC

This section provides details for each class and property defined by The Explainer Ontology.

Classes

Activation Clustersc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Activation_Clusters

has super-classes
Explainability Technique c

ALEc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ALE

Accumulated Local Effects
has super-classes
Influence Function c

Anchorc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Anchor

The anchors method explains individual predictions of any black box classification model by finding a decision rule that “anchors” the prediction sufficiently. A rule anchors a prediction if changes in other feature values do not affect the prediction.
Source
https://christophm.github.io/interpretable-ml-book/anchors.html
has super-classes
Simplification By Rule Extraction c

Anchor Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Anchor_Explanation

has super-classes
Feature Influence Explanation c

Architecture Modificationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Architecture_Modification

has super-classes
Explainability Technique c
has sub-classes
Attention Network c, Layer Modification c, Loss Modification c, Model Combination c

Attention Networkc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Attention_Network

has super-classes
Architecture Modification c

Audio Modalityc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#AudioModality

An audio modality in which an Explanation exists.
has super-classes
explanation modality

Caption Generationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Caption_Generation

has super-classes
Data-driven c

Categoricalc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Categorical

Categorical data type.
has super-classes
Data Type c

Compositec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Composite

has super-classes
Explainability Technique c

Computational Complexityc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ComputationalComplexity

Details relating the resources required by an Explainability Technique to generate an Explanation.
has sub-classes
Time Complexity c
is in range of
has complexity op

Conditional Plotsc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Conditional_Plots

has super-classes
Statistics c

Contrasting Feature Importance Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Contrasting_Feature_Importance_Explanation

has super-classes
Feature Influence Explanation c

Contrasting Gradient Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Contrasting_Gradient_Technique

has super-classes
Gradient-based Technique c

Contribution Distribution Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Contribution_Distribution_Explanation

has super-classes
Feature Influence Explanation c

Data Typec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#DataType

The types of model features (e.g., categorical, numerical, ordinal, image) that a given Explainability Technique can be used.
has sub-classes
Categorical c, Image c, Numerical c, Ordinal c
has members
Categorical ni, Image ni, Numerical ni, Ordinal ni

Data-drivenc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Data-driven

has super-classes
Explainability Technique c
has sub-classes
Caption Generation c, DisCERN c

DeepLIFTc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#DeepLIFT

has super-classes
Gradient-based Technique c

DiCEc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#DiCE

has super-classes
Optimisation Based c

DisCERNc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#DisCERN

has super-classes
Data-driven c

Explainability Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ExplainabilityTechnique

The technique used by an Explainer to generate an Explanation, such as rule-based learner, Decision Tree, Influence functions, Sensitivity, Saliency, and so on as categorised by Arrieta et. al. (2020)
has sub-classes
Activation Clusters c, Architecture Modification c, Composite c, Data-driven c, Feature Relevance c, Filter c, Knowledge Extraction c, Optimisation Based c, Probabilistic c, Simplification c, Statistics c
is in domain of
applicable problem type op, has applicable method types op, has complexity op, has concurrentness op, has explanation scope op, has output type op, has portability op, has presentation op, target type op
is in range of
utilises op

Explainerc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Explainer

An Explainer uses an Explainability Technique to generate Explanations for an AI Models output.
has super-classes
is in domain of
utilises op

Explainer Concurrentnessc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ExplainerConcurrentness

Captured if the technique is ante-hoc (i.e. use the same model for the task and creating explanations) or post-hoc (i.e. different models are used to create the eo:SystemRecommendation and eo:Explanation)
is equivalent to
{ Ante-hoc , Post-hoc }
is in range of
has concurrentness op
has members
Ante-hoc ni, Post-hoc ni

Explanation Scopec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ExplanationScope

The extent to which explanations generated by an explanation technique can be generalised.
is equivalent to
{ Cohort , Global , Local }
is in range of
has explanation scope op
has members
Cohort ni, Global ni, Local ni

Explanation Targetc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ExplanationTarget

The target of the Explanation Technique, i.e. the thing an Explanation is being generated for.
is in range of
target type op
has members
Data ni, Model ni, Prediction ni

Feature Influence Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Feature_Influence_Explanation

has super-classes
explanation
has sub-classes
Anchor Explanation c, Contrasting Feature Importance Explanation c, Contribution Distribution Explanation c, Saliency Map c, Sensitivity Map c

Feature Relevancec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Feature_Relevance

has super-classes
Explainability Technique c
has sub-classes
Game Theory Technique c, Gradient-based Technique c, Influence Function c, Saliency c

Filterc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Filter

has super-classes
Explainability Technique c

Game Theory Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Game_Theory_Technique

has super-classes
Feature Relevance c
has sub-classes
SHAP c

GradCam Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#GradCam_Technique

has super-classes
Gradient-based Technique c

Gradient-based Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Gradient-based_Technique

has super-classes
Feature Relevance c
has sub-classes
Contrasting Gradient Technique c, DeepLIFT c, GradCam Technique c, Integrated Gradient Technique c, SmoothGrad Technique c

Hidden-layer Clusteringc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Hidden-layer_Clustering

has super-classes
Saliency c

Imagec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Image

Image data type
has super-classes
Data Type c

Implementation Frameworkc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Implementation_Framework

Software library that provides a fundamental structure to support the development of applications for a specific environment.
has members
Any ni, LightGBM ni, PyTorch ni, Sklearn ni, TensorFlow 1 ni, TensorFlow 2 ni, XGBoost ni

Individual Condition Expectation Plotc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Individual_Condition_Expectation_Plot

Individual Conditional Expectation (ICE) plot gives a graphical depiction of the marginal effect of a variable on the response.
has super-classes
Influence Function c

Influence Functionc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Influence_Function

has super-classes
Feature Relevance c
has sub-classes
ALE c, Individual Condition Expectation Plot c, Partial Dependence Plot c

Integrated Gradient Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Integrated_Gradient_Technique

has super-classes
Gradient-based Technique c

Introspective Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Introspective_Explanation

has super-classes
contextual explanation

Knowledge Extractionc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Knowledge_Extraction

has super-classes
Explainability Technique c

Layer Modificationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Layer_Modification

has super-classes
Architecture Modification c

LIMEc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#LIME

LIME, or Local Interpretable Model-Agnostic Explanations, is an algorithm that can explain the predictions of any classifier or regressor in a faithful way, by approximating it locally with an interpretable model.
has super-classes
Simplification By Linear Regression c

Loss Modificationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Loss_Modification

has super-classes
Architecture Modification c

Model Combinationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Model_Combination

has super-classes
Architecture Modification c

Neighbourhood Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Neighbourhood_Explanation

has super-classes
explanation

Numericalc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Numerical

Numerical data type.
has super-classes
Data Type c

Optimisation Basedc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Optimisation_Based

has super-classes
Explainability Technique c
has sub-classes
DiCE c, Wachter c

Ordinalc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Ordinal

Ordinal data type.
has super-classes
Data Type c

Partial Dependence Plotc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Partial_Dependence_Plot

has super-classes
Influence Function c

Portabilityc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Portability

The type of model an Explainability Technique can be applied to.
is in range of
has portability op
has members
Model-agnostic ni, Model-class specific ni, Model-specific ni

Probabilisticc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Probabilistic

has super-classes
Explainability Technique c

Prototype Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Prototype_Explanation

has super-classes
explanation

Rationalisation Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Rationalisation_Explanation

has super-classes
contextual explanation

Saliencyc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Saliency

has super-classes
Feature Relevance c
has sub-classes
Hidden-layer Clustering c

Saliency Mapc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Saliency_Map

has super-classes
Feature Influence Explanation c

Semi-factual Explanationc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Semi-factual_Explanation

has super-classes
explanation

Sensitivity Mapc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Sensitivity_Map

has super-classes
Feature Influence Explanation c

SHAPc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SHAP

has super-classes
Game Theory Technique c

Simplication By Weights Dropoutc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplicationByWeightsDropout

has super-classes
Simplification c

Simplification By Decision Treec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationByDecisionTree

has super-classes
Simplification c

Simplification By Forestsc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationByForests

has super-classes
Simplification c

Simplification By kNNc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationBykNN

has super-classes
Simplification c

Simplification By Linear Proxy Modelc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationByLinear_Proxy_Model

has super-classes
Simplification c

Simplification By Linear Regressionc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationByLinear_Regression

has super-classes
Simplification c
has sub-classes
LIME c

Simplification By Rule Extractionc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SimplificationByRule_Extraction

has super-classes
Simplification c
has sub-classes
Anchor c

SmoothGrad Techniquec back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#SmoothGrad_Technique

has super-classes
Gradient-based Technique c

Statisticsc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Statistics

has super-classes
Explainability Technique c
has sub-classes
Conditional Plots c

Time Complexityc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Time_Complexity

Time complexity is the computational complexity that describes the amount of computer time it takes to run an algorithm.
Source
https://en.wikipedia.org/wiki/Time_complexity
has super-classes
Computational Complexity c
has members
Constant time ni, Exponential time ni, Factorial time ni, Linearithmic time ni, Log-logarithmic time ni, Logarithmic time ni, Quadratic time ni

Visual Modalityc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#VisualModality

A visual form in which an Explanation exists
has super-classes
explanation modality

Wachterc back to ToC or Class ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Wachter

Counterfactual explanation by Wachter et al.
Source
https://arxiv.org/pdf/1711.00399.pdf
has super-classes
Optimisation Based c

Object Properties

applicable problem typeop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#applicableProblemType

Relationship between an ExplainabilityTechnique and the type of eo:AI Task the technique can be used with. The Explanation Ontology defines three subclasses of eo:AI Task that the system was undertaking, namely for inductive, abductive, and deductive tasks.
has domain
Explainability Technique c
has range
a i task

has applicable method typesop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasApplicableMethodType

For ExplainabilityTechniques with Portability of model-class specific (works with a particular family of models) and model-specific, this property links to the relevant class in the eo:AI Method hierarchy that this ExplainabilityTechnique can be used with.
has domain
Explainability Technique c
has range
artificial intelligence method

has backendop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasBackend

The implementation backend of an AI Model or Explainer.

has complexityop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasComplexity

Relationahip between an ExplainabilityTechnique and some representation of the ComputationalComplexity of using it.
has domain
Explainability Technique c
has range
Computational Complexity c

has concurrentnessop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasConcurrentness

Relationship between an ExplinabilityTechnique and whether it is ante-hoc or post-hoc.

has characteristics: functional

has domain
Explainability Technique c
has range
Explainer Concurrentness c

has explanation scopeop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasExplanationScope

Relationship between an ExplinabilityTechnique and the extent to which explanations generated by it can be generalised, as defined in individuals of type ExplanationScope (e.g. local, cohort, global).
has domain
Explainability Technique c
has range
Explanation Scope c

has output typeop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasOutputType

Relationship between an ExplinabilityTechnique and the type of explanations generated by it, as defined by subclasses of eo:Explanation.
has domain
Explainability Technique c
has range
explanation

has portabilityop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasPortability

Link between Explainability Technique and the type of AI models it can be applied in terms of model-agnostic (works with any model family), model-class specific (works with a particular family of models), and model-specific as represented in individuals of type Portability.
has domain
Explainability Technique c
has range
Portability c

has presentationop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#hasPresentation

Link between Explainability Technique and the types of Information Content Entity (e.g. mathematical, language, computational) that is used to present the explanation.
has domain
Explainability Technique c
has range
s i o 000015

is compatible with feature typesop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#isCompatibleWithFeatureTypes

Link between Explainability Technique and the types of features used by the model being explained that the technique can be used with.

target typeop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#targetType

Link between Explainability Technique and target of the explaination - i.e. the thing being explained.
has domain
Explainability Technique c
has range
Explanation Target c

utilisesop back to ToC or Object Property ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#utilises

Link between an Explainer and the Explainability Technique that is uses.
has domain
Explainer c
has range
Explainability Technique c

Named Individuals

Ante-hocni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ante-hoc

An ante-hoc Explanation Technique use the same model for the task and creating explanations
belongs to
Explainer Concurrentness c

Anyni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Any

belongs to
Implementation Framework c

Categoricalni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#categorical

Data having a categorical type.
belongs to
Data Type c

Cohortni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#cohort

Explanations generated by an Explanation Technique with local scope explain a subgroup in a dataset or subspace in a model's decision space.
belongs to
Explanation Scope c

Constant timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Constant_time

O(1)
belongs to
Time Complexity c

Datani back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#data

Raw or feature data that are the target of an explanation.
belongs to
Explanation Target c

Exponential timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Exponential_time

O(2^n)
belongs to
Time Complexity c

Factorial timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Factorial_time

O(n!)
belongs to
Time Complexity c

Globalni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#global

Explanations generated by an Explanation Technique with global scope explain an AI Model.
belongs to
Explanation Scope c

Imageni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#image

Data of type image.
belongs to
Data Type c
dataset type c

LightGBMni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#LightGBM

belongs to
Implementation Framework c

Linearithmic timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Linearithmic_time

O(n log n)
belongs to
Time Complexity c

Localni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#local

Explanations generated by an Explanation Technique with local scope explain a single data point or prediction.
belongs to
Explanation Scope c

Log-logarithmic timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Log-logarithmic_time

O(log log n)
belongs to
Time Complexity c

Logarithmic timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Logarithmic_time

O(log n)
belongs to
Time Complexity c

Modelni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#model

The AI Model is the target of an Explanation.
belongs to
Explanation Target c

Model-agnosticni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#model-agnostic

An Explanation Technique that works with any model family.
belongs to
Portability c

Model-class specificni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#modelClassSpecific

An Explanation Technique that works with a particular family of models.
belongs to
Portability c

Model-specificni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#modelSpecific

An Explanation Technique that works with a specific type of model.
belongs to
Portability c

Multivariateni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#multivariate

belongs to
dataset type c

Numericalni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#numerical

Data having a numerical type
belongs to
Data Type c

Ordinalni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#ordinal

Data having an ordinal type
belongs to
Data Type c

Post-hocni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#post-hoc

A post-hoc Explanation Technique uses different models to create the eo:SystemRecommendation and eo:Explanation to that which were used as part of the Task
belongs to
Explainer Concurrentness c

Predictionni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#prediction

The Predictions are the target of an Explanation.
belongs to
Explanation Target c

PyTorchni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#PyTorch

belongs to
Implementation Framework c

Quadratic timeni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Quadratic_time

O(n^2)
belongs to
Time Complexity c

Sklearnni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#Sklearn

belongs to
Implementation Framework c

TensorFlow 1ni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#TensorFlow1

belongs to
Implementation Framework c

TensorFlow 2ni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#TensorFlow2

belongs to
Implementation Framework c

TensorFlow 2ni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#TensowFlow2

Textni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#text

belongs to
dataset type c

Time seriesni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#time_series

belongs to
dataset type c

Univariateni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#univariate

belongs to
dataset type c

XGBoostni back to ToC or Named Individual ToC

IRI: http://www.w3id.org/iSeeOnto/explainer#XGBoost

belongs to
Implementation Framework c

Legend back to ToC

c: Classes
op: Object Properties
dp: Data Properties
ni: Named Individuals

References back to ToC

Add your references here. It is recommended to have them as a list.

Acknowledgments back to ToC

The authors would like to thank Silvio Peroni for developing LODE, a Live OWL Documentation Environment, which is used for representing the Cross Referencing Section of this document and Daniel Garijo for developing Widoco, the program used to create the template used in this documentation.