-
Notifications
You must be signed in to change notification settings - Fork 0
/
main.ttl
1332 lines (1092 loc) · 90 KB
/
main.ttl
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
@prefix airo: <https://w3id.org/airo#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix terms: <http://purl.org/dc/terms/> .
@prefix earl: <http://www.w3.org/ns/earl#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .
@prefix : <https://example.org/ex#> . # for new declarations
######################################
# Description of bias
######################################
:Bias rdf:type owl:Class ;
terms:source "ISO/IEC 24027:2020, 3.3.2"@en ;
rdfs:comment "systematic difference in treatment of certain objects, people, or groups in comparison to others."@en ;
rdfs:label "bias"@en .
:CognitiveBias rdf:type owl:Class ;
rdfs:subClassOf :Bias ;
terms:source "ISO/IEC 24027:2020, 3.3.4"@en ;
rdfs:comment "human bias that might impact the design and application of a system."@en ;
rdfs:label "human cognitive bias"@en .
:AutomationBias rdf:type owl:Class ;
rdfs:subClassOf :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 3.3.1"@en ;
rdfs:comment "type of human cognitive bias due to over-reliance on the recommendations of an AI system."@en ;
rdfs:label "automation bias"@en .
:ConfirmationBias rdf:type owl:Class ;
rdfs:subClassOf :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 3.3.4"@en ;
rdfs:comment "type of human cognitive bias that favours predictions of AI systems that confirm pre-existing beliefs or hypotheses."@en ;
rdfs:label "confirmation bias"@en .
:DataBias rdf:type owl:Class ;
terms:source "ISO/IEC 24027:2020, 3.3.7"@en ;
rdfs:comment "data properties that if unaddressed lead to AI systems that perform better or worse for different groups."@en ;
rdfs:label "data bias"@en .
:StatisticalBias rdf:type owl:Class ;
terms:source "ISO 20501:2019, 3.3.9"@en ;
rdfs:comment "type of consistent numerical offset in an estimate relative to the true underlying value, inherent to most estimates."@en ;
rdfs:label "statistical bias"@en .
:Trustworthiness rdf:type airo:Entity ;
terms:source "ISO 17068:2017, 3.18"@en ;
rdfs:comment "quality of being dependable and reliable."@en ;
rdfs:label "trustworthiness"@en .
:Group rdf:type owl:Class ;
terms:source "ISO/IEC 24027:2020, 3.3.8"@en ;
rdfs:comment "subset of objects in a domain that are linked because they have shared characteristics."@en ;
rdfs:label "group"@en .
# The following are discussed in ISO/IEC 24027:2020 but not listed in the 'Terms' section of that document
:SocietalBias rdf:type owl:Class ;
rdfs:subClassOf :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 7.2.9"@en ;
rdfs:comment "Societal bias occurs when one or more similar cognitive biases (conscious or unconscious) are being held by many individuals in society. This societal bias originates from society at large and could be closely related to other cognitive or statistical biases. It manifests as data available about society that reflects historical patterns. Societal bias can also be considered a type of data bias."@en ;
rdfs:label "societal bias"@en .
:SelectionBias rdf:type owl:Class ;
rdfs:subClassOf :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.2.1"@en ;
rdfs:comment "Selection bias occurs when data is not collected randomly from the intended population."@en ;
rdfs:label "selection bias"@en .
# Instance-level declarations of different types of societal bias.
:ReligionBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to a person's religion"@en ;
rdfs:label "Religious bias"@en .
:RacialBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to the race of an individual"@en ;
rdfs:label "Racial bias"@en .
:SexualOrientationBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to the sexual orientation of an individual"@en ;
rdfs:label "Sexual orientation bias"@en .
:EthnicBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to the ethnicity of an individual"@en ;
rdfs:label "Ethic bias"@en .
:PoliticalBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to political views"@en ;
rdfs:label "Political bias"@en .
:PhilosophicalBeliefBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to philosophical beliefs"@en ;
rdfs:label "Philophical belief bias"@en .
:TradeUnionMembershipBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to trade union membership"@en ;
rdfs:label "Bias due to trade union membership"@en .
:GeneticDataBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to genetic data"@en ;
rdfs:label "Genetic data bias"@en .
:BiometricDataUseBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to the use of biometric data"@en ;
rdfs:label "Biometric data use bias"@en .
:HealthDataBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to health data"@en ;
rdfs:label "Health data bias"@en .
# Added due to its importance; not listed in GDPR or ISO/IEC 24027:2020
:GenderBias rdf:type :SocietalBias ;
rdfs:comment "Bias related to gender"@en ;
rdfs:label "Gender bias"@en .
:SamplingBias rdf:type :SelectionBias ;
terms:source "ISO/IEC 24027:2020, 7.3.2.2"@en ;
rdfs:comment "Sampling bias occurs when data is not collected randomly from the intended population."@en ;
rdfs:label "sampling bias"@en .
:CoverageBias rdf:type :SelectionBias ;
terms:source "ISO/IEC 24027:2020, 7.3.2.3"@en ;
rdfs:comment "Coverage bias occurs when a population represented in a dataset does not match the population that the ML model is making predictions about."@en ;
rdfs:label "coverage bias"@en .
:NonResponseBias rdf:type :SelectionBias ;
terms:source "ISO/IEC 24027:2020, 7.3.2.4"@en ;
rdfs:comment "Non-response bias (also called participation bias) occurs when people from certain groups opt-out of surveys at different rates than users from other groups."@en ;
rdfs:label "non-response bias"@en .
:DataLabelBias rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.3"@en ;
rdfs:comment "The labelling process itself potentially introduces the cognitive or societal biases described in subclause 7.2 to the data."@en ;
rdfs:label "data labels and labelling process"@en .
:NonRepresentativeSampling rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.4"@en ;
rdfs:comment "If a dataset is not representative of the intended deployment environment, then the model has the potential to learn biases based on the ways in which the data is non-representative."@en ;
rdfs:label "non-representative sampling"@en .
:MissingFeaturesLabels rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.5"@en ;
rdfs:comment "Features are often missing from individual training samples. If the frequency of missing features is higher for one group than another then this presents another vector for bias."@en ;
rdfs:label "missing features and labels"@en .
:DataProcessing rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.6"@en ;
rdfs:comment "Bias might also creep in due to pre-processing (or post-processing) of data, even though the original data would not have led to any bias."@en ;
rdfs:label "data processing bias"@en .
:SimpsonsParadox rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.7"@en ;
rdfs:comment "Simpson's paradox manifests when a trend that is indicated in individual groups of data reverses when the groups of data are combined."@en ;
rdfs:label "Simpson's paradox"@en .
:ConfoundingVariableBias rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.8"@en ;
rdfs:coment "A confounding variable that influences both the dependent variable and independent variable causing a spurious association."@en ;
rdfs:label "confounding variables"@en .
:NonNormality rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.9"@en ;
rdfs:comment "if the data is subject to a different distribution [from normal] (e.g. Chi-Square, Beta, Lorentz, Cauchy, Weibull or Pareto) the result might be biased and misleading."@en ;
rdfs:label "non-normality"@en .
:OutlierNoiseBias rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.10"@en ;
rdfs:comment "Outliers are extreme data values that, if real, represent very low probability events of the to-be-modelled data. Noise is distortion which is characterised by a statistically-distributed variation of a physical quantity."@en ;
rdfs:label "outliers and noise"@en .
:DataAggregationBias rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.11; http://arxiv.org/abs/1901.10002"@en ;
rdfs:comment "Aggregating data covering different groups of objects that might have different statistical distributions can introduce bias into the data used to train AI systems"@en ;
rdfs:label "data aggregation bias"@en .
:DistributedTrainingBias rdf:type :DataBias ;
terms:source "ISO/IEC 24027:2020, 7.3.12"@en ;
rdfs:comment "Distributed ML might introduce its own cause for data bias as the different sources of data might not have the same distribution of feature space."@en ;
rdfs:label "distributed training bias"@en .
:GroupAttributionBias rdf:type :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 7.2.3"@en ;
rdfs:comment "Group attribution bias occurs when a human assumes that what is true for an individual or object is also true for everyone, or all objects, in that group."@en ;
rdfs:label "group attribution bias"@en .
:ImplicitBias rdf:type :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 7.2.4"@en ;
rdfs:comment "Implicit bias occurs when a human makes an association or assumption based on their mental models and memories."@en ;
rdfs:label "implicit bias"@en .
:InGroupBias rdf:type :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 7.2.6"@en ;
rdfs:comment "In-group bias occurs when showing partiality to one's own group or own characteristics."@en ;
rdfs:label "in-group bias"@en .
:OutGroupHomogeneityBias rdf:type :CognitiveBias ;
terms:source "ISO/IEC 24027:2020, 7.2.7"@en ;
rdfs:comment "Out-group homogeneity bias occurs when seeing out-group members as more alike than in-group members when comparing attitudes, values, personality traits, and other characteristics."@en ;
rdfs:label "out-group homogeneity bias"@en .
:WYSIATIBias rdf:type :CognitiveBias ;
terms:source "ISO IEC 24027:2020, 7.2.8 / https://www.microsoft.com/en-us/research/uploads/prod/2020/05/Fairlearn_WhitePaper-2020-09-22.pdf"@en ;
rdfs:comment "'What You See Is All There Is' (WYSIATI) bias occurs when a human looks for information that confirms their beliefs, overlooks contradicting information, and draws conclusions based on what is familiar."@en ;
rdfs:label "What you see is all there is bias"@en .
################################################################################
# Bias audits, a type of Documentation which must have an ML Model as domain
################################################################################
# Implicitly in being a subclass of airo:Documentation, ANY airo:AIComponent can have a bias audit on it
:BiasAudit rdf:type owl:Class ;
rdfs:subClassOf airo:Documentation ;
rdfs:comment "An audit which focuses on measuring bias in the AI system"@en ;
rdfs:label "Bias audit"@en .
# :hasAudit property no longer required; use airo:hasDocumentation with a :BiasAudit.
# Very important to store the date that the audit was conducted
:auditDate rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAudit ;
rdfs:range xsd:date ;
rdfs:comment "Indicates the date that a bias audit was conducted"@en ;
rdfs:label "Bias audit date" .
:BiasAudit rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :auditDate ;
owl:cardinality "1"^^xsd:integer
] .
# A particular bias assertion should consist of:
# 1. A description of the groups involved in the test, and the reason for this test.
# 2. Indication of the fairness metric used for the test.
# 3. Indication of the test result (e.g., a float value)
# 4. Indication of the threshold for passing / failure, justifying what this threshold is supposed to represent.
# 5. Indication of whether the test passed or failed.
# 6. Indication of biases implied by this test.
:BiasAssertion rdf:type owl:Class ;
rdfs:subClassOf earl:Assertion ;
rdfs:comment "Indicates that an assertion measures bias"@en ;
rdfs:label "Bias assertion"@en .
# Constraint to ensure any :BiasAssertion has a :Bias test subject
:BiasAssertion rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :testsBias ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] ,
[
rdf:type owl:Restriction ;
owl:onProperty :groupDescription ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] ,
[
rdf:type owl:Restriction ;
owl:onProperty :testThreshold ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
:BiasAssertion rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :hasFairnessMetric ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
# A custom 'tests bias' type. Purpose: Simplifies documentation.
# Avoids constructing specific conditions as in earl:TestCriterion ;
# 'Pass' when the bias tested for is deemed present, 'fail' when the bias tested for is deemed not present.
:testsBias rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAssertion ;
rdfs:range :Bias ;
rdfs:comment "Connects an Assertion with the bias type it is testing for."@en ;
rdfs:label "Tests bias"@en .
:hasAssertion rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAudit ;
rdfs:range earl:Assertion ;
rdfs:comment "Connects a bias audit with an assertion"@en ;
rdfs:label "has assertion"@en .
# A URI pointing to where the results reside
:testResultURI rdf:type owl:DatatypeProperty ;
rdfs:domain :BiasAssertion ;
rdfs:range xsd:anyURI ;
rdfs:comment "URI pointing to the test result (e.g., GitHub link)."@en ;
rdfs:label "Result URI"@en .
# Description of the groups being tested in a bias assertion
:groupDescription rdf:type owl:DatatypeProperty ;
rdfs:domain :BiasAssertion ;
rdfs:range xsd:string ;
rdfs:comment "Provides a description of the groups which are compared in a bias assertion"@en ;
rdfs:label "Group description"@en .
# Test thresholds
:Threshold rdf:type owl:Class ;
rdfs:comment "A threshold associated with a bias test"@en ;
rdfs:label "Threshold"@en .
:testThreshold rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAssertion ;
rdfs:range :Threshold;
rdfs:comment "Indicates the threshold associated with a particular bias test assertion"@en ;
rdfs:label "test threshold"@en .
:ThresholdRange rdf:type owl:Class ;
rdfs:subClassOf :Threshold ;
rdfs:comment "A threshold which indicates a range of acceptable or unacceptable values"@en ;
rdfs:label "Threshold range"@en .
:ThresholdBinary rdf:type owl:Class ;
rdfs:subClassOf :Threshold ;
rdfs:comment "A threshold which indicates a value, above or below which a test may be considered to have passed or failed"@en ;
rdfs:label "Threshold binary"@en .
:thresholdBegin rdf:type owl:DatatypeProperty ;
rdfs:domain :ThresholdRange ;
# rdfs:range xsd:float , xsd:integer ;
rdfs:range [
rdf:type rdfs:Datatype ;
owl:unionOf ( xsd:float xsd:integer )
] ;
rdfs:comment "Indicates a float or integer beginning value for a range threshold"@en ;
rdfs:label "threshold begin"@en .
:thresholdEnd rdf:type owl:DatatypeProperty ;
rdfs:domain :ThresholdRange ;
rdfs:range [
rdf:type rdfs:Datatype ;
owl:unionOf ( xsd:float xsd:integer )
] ;
rdfs:comment "Indicates a float or integer ending value for a range threshold"@en ;
rdfs:label "threshold end"@en .
:binaryThreshold rdf:type owl:DatatypeProperty ;
rdfs:domain :ThresholdBinary ;
rdfs:range [
rdf:type rdfs:Datatype ;
owl:unionOf ( xsd:float xsd:integer )
] ;
rdfs:comment "Indicates a float or integer value for the binary threshold"@en ;
rdfs:label "binary threshold"@en .
# OWL constraints on threshold objects
:ThresholdRange rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :thresholdBegin ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] ,
[
rdf:type owl:Restriction ;
owl:onProperty :thresholdEnd ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
:ThresholdBinary rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :binaryThreshold ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
# Test pass or fail indicates the nature of the thresholds
# Expand EARL to allow for numerical test results (not only pass / fail)
:FloatPass rdf:type owl:Class ;
rdfs:subClassOf earl:Pass ;
rdfs:comment "Indicates that a test results in a pass, with a floating point test result."@en ;
rdfs:label "Float pass"@en .
:FloatFail rdf:type owl:Class ;
rdfs:subClassOf earl:Fail ;
rdfs:comment "Indicates that a test results in a fail, with a floating point test result."@en ;
rdfs:label "Float fail"@en .
:IntegerPass rdf:type owl:Class ;
rdfs:subClassOf earl:Pass ;
rdfs:comment "Indicates that a test results in a pass, with an integer test result."@en ;
rdfs:label "Integer pass"@en .
:IntegerFail rdf:type owl:Class ;
rdfs:subClassOf earl:Fail ;
rdfs:comment "Indicates that a test results in a fail, with an integer test result."@en ;
rdfs:label "Integer fail"@en .
:floatValue rdf:type owl:DatatypeProperty ;
rdfs:domain :FloatPass , :FloatFail ;
rdfs:range xsd:float ;
rdfs:comment "Indicates the float value result associated with a FloatPass or FloatFail test result."@en ;
rdfs:label "float value"@en .
:integerValue rdf:type owl:DatatypeProperty ;
rdfs:domain :IntegerPass , :IntegerFail ;
rdfs:range xsd:integer ;
rdfs:comment "Indicates the integer value result associated with an IntegerPass or IntegerFail test result."@en ;
rdfs:label "integer value"@en .
# Enforce cardinality constraints for properties
:FloatPass rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :floatValue ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
:FloatFail rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :floatValue ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
:IntegerPass rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :integerValue ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
:IntegerFail rdfs:subClassOf [
rdf:type owl:Restriction ;
owl:onProperty :integerValue ;
owl:cardinality "1"^^xsd:nonNegativeInteger
] .
# Add possibility to describe bias metrics.
:FairnessMetric rdf:type owl:Class ;
terms:source "IBM AI Fairness 360 Glossary, https://aif360.res.ibm.com/resources#glossary"@en ;
rdfs:comment "A quantification of unwanted bias in training data or models"@en ;
rdfs:label "Fairness metric"@en .
:hasFairnessMetric rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAssertion ;
rdfs:range :FairnessMetric ;
rdfs:comment "Indicates that a test uses a particular fairness metric."@en ;
rdfs:label "has fairness metric"@en .
# Fairness metrics from IBM AIF360
:StatisticalParityDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-stat-parity-diff.html?context=cpdaas"@en ;
rdfs:comment "The Statistical parity difference metric calculates the difference in the ratio of favorable outcomes between monitored groups and reference groups"@en ;
rdfs:label "Statistical parity difference"@en .
:DisparateImpact rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-disparate-impact.html?context=analytic"@en ;
rdfs:comment "Disparate impact compares the percentage of favorable outcomes for a monitored group to the percentage of favorable outcomes for a reference group"@en ;
rdfs:label "Disparate impact"@en .
:EqualOpportunityDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Web Page, https://aif360.res.ibm.com/"@en ;
rdfs:comment "The difference of true positive rates between the unprivileged and the privileged groups"@en ;
rdfs:label "Equal opportunity difference"@en .
:AverageOddsDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-fairness-metrics-ovr.html?context=cpdaas"@en ;
rdfs:comment "The average odds difference gives the percentage of transactions that was incorrectly scored by your model"@en ;
rdfs:label "Average odds difference"@en .
:AverageAbsOddsDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-fairness-metrics-ovr.html?context=cpdaas"@en ;
rdfs:comment "The average absolute odds difference compares the average of absolute difference in false positive rates and true positive rates between monitored groups and reference groups"@en ;
rdfs:label "Average absolute odds difference"@en .
:ErrorRateDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-fairness-metrics-ovr.html?context=cpdaas"@en ;
rdfs:comment "The error rate difference calculates the percentage of transactions that are incorrectly scored by your model"@en ;
rdfs:label "Error rate difference"@en .
:ErrorRateRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric"@en ;
rdfs:comment "(False positives + False negatives) divided by (Positives + Negatives) "@en ; # (FP+FN)/(P+N)
rdfs:label "Error rate ratio"@en .
:FalseDiscoveryRateDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-fairness-metrics-ovr.html?context=cpdaas"@en ;
rdfs:comment "The false discovery rate difference gives the amount of false positive transactions as a percentage of all transactions with a positive outcome"@en ;
rdfs:label "False discovery rate difference"@en .
:FalseDiscoveryRateRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric"@en ;
rdfs:comment "(False discovery rate for unprivileged group) divided by (False discovery rate for privileged group)"@en ; # FDRu/FDRp
rdfs:label "False discovery rate ratio"@en .
:FalseNegativeRateRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric"@en ;
rdfs:comment "(False negative rate for the unprivileged group) divided by (False negative rate for the privileged group)"@en ; # FNRu/FNRp
rdfs:label "False negative rate ratio"@en .
:FalseOmissionRateDifference rdf:type :FairnessMetric ;
terms:source "IBM Cloud Pak for Data Docs, https://dataplatform.cloud.ibm.com/docs/content/wsj/model/wos-fairness-metrics-ovr.html?context=cpdaas"@en ;
rdfs:comment "The false omission rate difference gives the number of false negative transactions as a percentage of all transactions with a negative outcome"@en ;
rdfs:label "False omission rate difference"@en .
:FalseOmissionRateRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric"@en ;
rdfs:comment "(False omission rate for unprivileged group) divided by (False omission rate for privileged group)"@en ; # FORu/FORp
rdfs:label "False omission rate ratio"@en .
:FalsePositiveRateRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "(False positive rate for unprivileged group) divided by (False positive rate for privileged group)"@en ;
rdfs:label "False positive rate ratio"@en .
:FalseNegativeRateDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "(False negative rate for unprivileged group) divided by (False negative rate for privileged group)"@en ; # FNRu-FNRp
rdfs:label "False negative rate difference"@en .
:FalsePositiveRateDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "(False positive rate for unprivileged group) minus (False positive rate for privileged group)"@en ; # FPRu-FPRp
rdfs:label "False positive rate difference"@en .
:TruePositiveRateDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The ratio of true positives to positive examples in the dataset, TPR = TP/P"@en ;
rdfs:label "True positive rate difference"@en .
:CoefficientOfVariation rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The square root of two times the generalize entropy index"@en ;
rdfs:label "Coefficient of variation"@en .
:TheilIndex rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The Theil index is the generalized entropy index with alpha = 1" ;
rdfs:label "Theil index"@en .
:BetweenGroupTheilIndex rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The between group generalized entropy index with alpha = 1"@en ;
rdfs:label "Between group Theil index"@en .
:BetweenGroupCoefficientOfVariation rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The square root of two times the between group generalized entropy index with alpha = 2"@en ;
rdfs:label "Between group coefficient of variation"@en .
:BetweenAllGroupsCoefficientOfVariation rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "The square root of two times the between all groups generalized entropy index with alpha = 2"@en ;
rdfs:label "Between all groups coefficient of variation"@en .
:BetweenAllGroupsGeneralisedEntropyIndex rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "Between-group generalized entropy index that uses all combinations of groups based on self.dataset.protected_attributes"@en ;
rdfs:label "Between all groups generalised entropy index"@en .
:ConsistencyMetric rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "Individual fairness metric that measures how similar the labels are for similar instances"@en ;
rdfs:label "Consistency metric"@en .
:GeneralisedEntropyIndex rdf:type :FairnessMetric ;
terms:source "A Unified Approach to Quantifying Algorithmic Unfairness: Measuring Individual & Group Unfairness via Inequality Indices. https://arxiv.org/abs/1807.00787" ;
rdfs:comment "Generalized entropy index is a measure of unfairness which can be interpreted as a measure of information theoretic redundancy in data."@en ;
rdfs:label "Generalised entropy index"@en .
:BetweenGroupGeneralisedEntropyIndex rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.ClassificationMetric.html#aif360.metrics.ClassificationMetric" ;
rdfs:comment "Between-group generalized entropy index that uses self.privileged_groups and self.unprivileged_groups as the only two groups"@en ;
rdfs:label "Between groups generalised entropy index"@en .
:MeanEuclideanDistanceDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Difference of the average Euclidean distances"@en ;
rdfs:label "Mean Euclidean distance difference"@en .
:MeanEuclideanDistanceRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Ratio of the average Euclidean distances"@en ;
rdfs:label "Mean Euclidean distance ratio"@en .
:MeanMahalanobisDistanceDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Difference of the average Mahalanobis distances"@en ;
rdfs:label "Mean Mahalanobis distance difference"@en .
:MeanMahalanobisDistanceRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Ratio of the average Mahalanobis distances"@en ;
rdfs:label "Mean Mahalanobis distance ratio"@en .
:MeanManhattanDistanceDifference rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Difference of the average Manhattan distances"@en ;
rdfs:label "Mean Manhattan distance difference"@en .
:MeanManhattanDistanceRatio rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.SampleDistortionMetric.html#aif360.metrics.SampleDistortionMetric"@en ;
rdfs:comment "Ratio of the average Manhattan distances"@en ;
rdfs:label "Mean Manhattan distance ratio"@en .
:BaseRateMetric rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.BinaryLabelDatasetMetric.html#aif360.metrics.BinaryLabelDatasetMetric"@en ;
rdfs:comment "Pr(Y=1) = P/(P+N)"@en ;
rdfs:label "Base rate metric"@en .
:SmoothedEmpiricalDifferentialFairness rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.BinaryLabelDatasetMetric.html#aif360.metrics.BinaryLabelDatasetMetric"@en ;
rdfs:comment "Smoothed empirical differential fairness"@en ;
rdfs:label "Smoothed empirical differential fairness"@en .
:RichSubgroupMetric rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.BinaryLabelDatasetMetric.html#aif360.metrics.BinaryLabelDatasetMetric"@en ;
rdfs:comment "Audit dataset with respect to rich subgroups defined by linear thresholds of sensitive attributes"@en ;
rdfs:label "Rich subgroup metric"@en .
:DifferentialFairnessBiasAmplification rdf:type :FairnessMetric ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.MDSSClassificationMetric.html#aif360.metrics.MDSSClassificationMetric"@en ;
rdfs:comment "Bias amplification is the difference in smoothed EDF between the classifier and the original dataset"@en ;
rdfs:label "Differential fairness bias amplification"@en .
:MDSSBiasScore rdf:type :FairnessMetric ; # from score_groups of MDSSClassificationMetric
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/generated/aif360.metrics.MDSSClassificationMetric.html#aif360.metrics.MDSSClassificationMetric"@en ;
rdfs:comment "The bias score for a prespecified group of records"@en ;
rdfs:label "MDSS bias score"@en .
# Mitigation applied on tests
:mitigationApplied rdf:type owl:ObjectProperty ;
rdfs:domain :BiasAssertion , :BiasAudit ;
rdfs:range :BiasMitigationMeasure ;
rdfs:comment "Indicates a mitigation measure applied to a bias assertion"@en ;
rdfs:label "Mitigation applied"@en .
:BiasMitigationMeasure rdf:type owl:Class ;
rdfs:subClassOf airo:Control ;
rdfs:comment "A mitigation measure for bias"@en ;
rdfs:label "Bias mitigation measure"@en .
# Bias mitigation metrics in IBM AI fairness 360
:OptimizedPreprocessing rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "A preprocessing technique that learns a probabilistic transformation that edits the features and labels in the data with group fairness, individual distortion, and data fidelity constraints and objectives."@en ;
rdfs:label "Optimized preprocessing"@en .
:DisparateImpactRemover rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Disparate impact remover is a preprocessing technique that edits feature values increase group fairness while preserving rank-ordering within groups."@en ;
rdfs:label "Disparate impact remover"@en .
:EqualizedOddsPostprocessing rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "A post-processing technique that solves a linear program to find probabilities with which to change output labels to optimize equalized odds."@en ;
rdfs:label "Equalized odds postprocessing"@en .
:Reweighing rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Reweighing is a preprocessing technique that Weights the examples in each (group, label) combination differently to ensure fairness before classification."@en ;
rdfs:label "Reweighing"@en .
:RejectOptionClassification rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Reject option classification is a postprocessing technique that gives favorable outcomes to unprivileged groups and unfavorable outcomes privileged groups in a confidence band around the decision boundary with the highest uncertainty."@en ;
rdfs:label "Reject option classification"@en .
:PrejudiceRemoverRegularizer rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Prejudice remover is an in-processing technique that adds a discrimination-aware regularization terms to the learning objective."@en ;
rdfs:label "Prejudice remover regularizer"@en .
:CalibratedEqualizedOddsPostprocessing rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Calibrated equalized odds postprocessing is a post-processing technique that optimizes over calibrated classifier score outputs to find probabilities with which to change output labels with an equalized odds objective."@en ;
rdfs:label "Calibrated equalized odds postprocessing"@en .
:LearningFairRepresentations rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Learning fair representations is a pre-processing technique that finds a latent representation which encodes the data well but obfuscates information about protected attributes."@en ;
rdfs:label "Learning fair representations"@en .
:AdversarialDebiasing rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Adversarial debiasing is an in-processing technique that learns a classifier to maximize prediction accuracy and simultaneously reduce an adversary's ability to determine the protected attribute from the predictions."@en ;
rdfs:label "Adversarial debiasing"@en .
:MetaAlgorithmFairClassification rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "The meta algorithm here takes the fairness metric as part of the input and returns a classifier optimized w.r.t."@en ;
rdfs:label "Meta algorithm fair classification"@en .
:RichSubgroupFairness rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Model is an algorithm for learning classifiers that are fair with respect to rich subgroups"@en ;
rdfs:label "Rich subgroup fairness"@en .
:ExponentiatedGradientReduction rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Exponentiated gradient reduction for fair classification"@en ;
rdfs:label "Exponentiated gradient reduction"@en .
:GridSearchReduction rdf:type :BiasMitigationMeasure ;
terms:source "IBM AIF360 Python Docs, https://aif360.readthedocs.io/en/stable/modules/algorithms.html"@en ;
rdfs:comment "Grid search reduction for fair classification or regression"@en ;
rdfs:label "Grid search reduction"@en .
# Expand with AIO categories --- should the AIO namespace be used?
:ComputationalBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=root"@en ;
rdfs:comment "A systematic tendency which causes differences between results and facts. The bias exists in numbers of the process of data analysis, including the source of the data, the estimator chosen, and the ways the data was analyzed."@en ;
rdfs:label "computational bias"@en ;
rdfs:subClassOf :Bias .
:ProcessingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FProcessingBias"@en ;
rdfs:comment "Judgement modulated by affect, which is influenced by the level of efficacy and efficiency in information processing; in cognitive sciences, processing bias is often referred to as an aesthetic judgement."@en ;
rdfs:label "processing bias"@en ;
rdfs:subClassOf :ComputationalBias .
:AmplificationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FAmplificationBias"@en ;
rdfs:comment "Arises when the distribution over prediction outputs is skewed in comparison to the prior distribution of the prediction target"@en ;
rdfs:label "amplification bias"@en ;
rdfs:subClassOf :ProcessingBias .
:ErrorPropagationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FErrorPropagationBias"@en ;
rdfs:comment "The effect of variables' uncertainties (or errors, more specifically random errors) on the uncertainty of a function based on them"@en ;
rdfs:label "error propagation bias"@en ;
rdfs:subClassOf :ProcessingBias .
:InheritedBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FInheritedBias"@en ;
rdfs:comment "Arises when applications that are built with machine learning are used to generate inputs for other machine learning algorithms. If the output is biased in any way, this bias may be inherited by systems using the output as input to learn other models."@en ;
rdfs:label "inherited bias"@en ;
rdfs:subClassOf :ProcessingBias .
:ModelSelectionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FModelSelectionBias"@en ;
rdfs:comment "The bias introduced while using the data to select a single seemingly 'best' model from a large set of models employing many predictor variables. Model selection bias also occurs when an explanatory variable has a weak relationship with the response variable."@en ;
rdfs:label "model selection bias"@en ;
rdfs:subClassOf :ComputationalBias .
:SurvivorshipBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FSurvivorshipBias"@en ;
rdfs:comment "Tendency for people to focus on the items, observations, or people that 'survive' or make it past a selection process, while overlooking those that did not."@en ;
rdfs:label "survivorship bias"@en ;
rdfs:subClassOf :ComputationalBias .
:DataGenerationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FDataGenerationBias"@en ;
rdfs:comment "Arises from the addition of synthetic or redundant data samples to a dataset"@en ;
rdfs:label "data generation bias"@en ;
rdfs:subClassOf :DataBias .
:DetectionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FDetectionBias"@en ;
rdfs:comment "Systematic differences between groups in how outcomes are determined and may cause an over- or underestimation of the size of the effect."@en ;
rdfs:label "detection bias"@en ;
rdfs:subClassOf :DataBias .
:EcologicalFallacyBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FEcologicalFallacyBias"@en ;
rdfs:comment "Occurs when an inference is made about an individual based on their membership within a group."@en ;
rdfs:label "ecological fallacy bias"@en ;
rdfs:subClassOf :DataBias .
:EvaluationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FEvaluationBias"@en ;
rdfs:comment "Arises when the testing or external benchmark populations do not equally represent the various parts of the user population or from the use of performance metrics that are not appropriate for the way in which the model will be used."@en ;
rdfs:label "evaluation bias"@en ;
rdfs:subClassOf :DataBias .
:ExclusionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FExclusionBias"@en ;
rdfs:comment "When specific groups of user populations are excluded from testing and subsequent analyses."@en ;
rdfs:label "exclusion bias"@en ;
rdfs:subClassOf :DataBias .
:MeasurementBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FMeasurementBias"@en ;
rdfs:comment "Arises when features and labels are proxies for desired quantities, potentially leaving out important factors or introducing group or input-dependent noise that leads to differential performance."@en ;
rdfs:label "measurement bias"@en ;
rdfs:subClassOf :DataBias .
:PopularityBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FPopularityBias"@en ;
rdfs:comment "A form of selection bias that occurs when items that are more popular are more exposed and less popular items are under-represented."@en ;
rdfs:label "popularity bias"@en ;
rdfs:subClassOf :DataBias .
:PopulationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FPopulationBias"@en ;
rdfs:comment "A form of selection bias that occurs when items that are more popular are more exposed and less popular items are under-represented.aSystematic distortions in demographics or other user characteristics between a population of users represented in a dataset or on a platform and some target population."@en ;
rdfs:label "population bias"@en ;
rdfs:subClassOf :DataBias .
:RepresentationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FRepresentationBias"@en ;
rdfs:comment "Arises due to non-random sampling of subgroups, causing trends estimated for one population to not be generalizable to data collected from a new population."@en ;
rdfs:label "representation bias"@en ;
rdfs:subClassOf :DataBias .
:TemporalBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FTemporalBias"@en ;
rdfs:comment "Bias that arises from differences in populations and behaviors over time."@en ;
rdfs:label "temporal bias"@en ;
rdfs:subClassOf :DataBias .
:UncertaintyBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FUncertaintyBias"@en ;
rdfs:comment "Arises when predictive algorithms favor groups that are better represented in the training data, since there will be less uncertainty associated with those predictions."@en ;
rdfs:label "uncertainty bias"@en ;
rdfs:subClassOf :DataBias .
:UseAndInterpretationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FUseAndInterpretationBias"@en ;
rdfs:comment "An information-processing bias, the tendency to inappropriately analyze ambiguous stimuli, scenarios and events."@en ;
rdfs:label "use and interpretation bias"@en ;
rdfs:subClassOf :Bias .
:ActivityBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FActivityBias"@en ;
rdfs:comment "A type of selection bias that occurs when systems/platforms get their training data from their most active users, rather than those less active (or inactive)."@en ;
rdfs:label "activity bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:ConceptDriftBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FConceptDriftBias"@en ;
rdfs:comment "Use of a system outside the planned domain of application, and a common cause of performance gaps between laboratory settings and the real world."@en ;
rdfs:label "concept drift bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:ContentProductionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FContentProductionBias"@en ;
rdfs:comment "Arises from structural, lexical, semantic, and syntactic differences in the contents generated by users."@en ;
rdfs:label "content production bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:DataDredgingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FDataDredgingBias"@en ;
rdfs:comment "A statistical bias in which testing huge numbers of hypotheses of a dataset may appear to yield statistical significance even when the results are statistically nonsignificant."@en ;
rdfs:label "data dredging bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:EmergentBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FEmergentBias"@en ;
rdfs:comment "Emergent bias is the result of the use and reliance on algorithms across new or unanticipated contexts."@en ;
rdfs:label "emergent bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:FeedbackLoopBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FFeedbackLoopBias"@en ;
rdfs:comment "Effects that may occur when an algorithm learns from user behavior and feeds that behavior back into the model."@en ;
rdfs:label "feedback loop bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:HostileAttributionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FHostileAttributionBias"@en ;
rdfs:comment "A bias wherein individuals perceive benign or ambiguous behaviors as hostile."@en ;
rdfs:label "hostile attribution bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:LinkingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FLinkingBias"@en ;
rdfs:comment "Arises when network attributes obtained from user connections, activities, or interactions differ and misrepresent the true behavior of the users."@en ;
rdfs:label "linking bias"@en ;
rdfs:subClassOf :UseAndInterpretationBias .
:GroupBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FGroupBias"@en ;
rdfs:comment "A pattern of favoring members of one's in-group over out-group members. This can be expressed in evaluation of others, in allocation of resources, and in many other ways."@en ;
rdfs:label "group bias"@en ;
rdfs:subClassOf :CognitiveBias .
:DeploymentBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FDeploymentBias"@en ;
rdfs:comment "Arises when systems are used as decision aids for humans, since the human intermediary may act on predictions in ways that are typically not modeled in the system. However, it is still individuals using the deployed system."@en ;
rdfs:label "deployment bias"@en ;
rdfs:subClassOf :GroupBias .
:FundingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FFundingBias"@en ;
rdfs:comment "Arises when biased results are reported in order to support or satisfy the funding agency or financial supporter of the research study, but it can also be the individual researcher."@en ;
rdfs:label "funding bias"@en ;
rdfs:subClassOf :GroupBias .
:GroupthinkBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FGroupthinkBias"@en ;
rdfs:comment "A psychological phenomenon that occurs when people in a group tend to make non-optimal decisions based on their desire to conform to the group, or fear of dissenting with the group. In groupthink, individuals often refrain from expressing their personal disagreement with the group, hesitating to voice opinions that do not align with the group."@en ;
rdfs:label "Groupthink bias"@en ;
rdfs:subClassOf :GroupBias .
:IndividualBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FIndividualBias"@en ;
rdfs:comment "Individual bias is a persistent point of view or limited list of such points of view that one applies ('parent', 'academic', 'professional', or etc.)."@en ;
rdfs:label "individual bias"@en ;
rdfs:subClassOf :CognitiveBias .
:AnchoringBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FAnchoringBias"@en ;
rdfs:comment "A cognitive bias, the influence of a particular reference point or anchor on people’s decisions. Often more fully referred to as anchoring-and-adjustment, or anchoring-and-adjusting: after an anchor is set, people adjust insufficiently from that anchor point to arrive at a final answer. Decision makers are biased towards an initially presented value."@en ;
rdfs:label "anchoring bias"@en ;
rdfs:subClassOf :IndividualBias .
:RankingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FRankingBias"@en ;
rdfs:comment "The idea that top-ranked results are the most relevant and important and will result in more clicks than other results."@en ;
rdfs:label "ranking bias"@en ;
rdfs:subClassOf :AnchoringBias .
:AnnotatorReportingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FAnnotatorReportingBias"@en ;
rdfs:comment "When users rely on automation as a heuristic replacement for their own information seeking and processing. A form of individual bias but often discussed as a group bias, or the larger effects on natural language processing models."@en ;
rdfs:label "annotator reporting bias"@en ;
rdfs:subClassOf :IndividualBias .
:AvailabilityHeuristicBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FAvailabilityHeuristicBias"@en ;
rdfs:comment "A mental shortcut whereby people tend to overweight what comes easily or quickly to mind, meaning that what is easier to recall—e.g., more “available”—receives greater emphasis in judgement and decision-making."@en ;
rdfs:label "availability heuristic bias"@en ;
rdfs:subClassOf :IndividualBias .
:BehaviouralBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FBehavioralBias"@en ;
rdfs:comment "Systematic distortions in user behavior across platforms or contexts, or across users represented in different datasets."@en ;
rdfs:label "behavioural bias"@en ;
rdfs:subClassOf :IndividualBias .
:DunningKrugerEffectBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FDunning-KrugerEffectBias"@en ;
rdfs:comment "The tendency of people with low ability in a given area or task to overestimate their self-assessed ability. Typically measured by comparing self-assessment with objective performance, often called subjective ability and objective ability, respectively."@en ;
rdfs:label "Dunning-Kruger effect bias"@en ;
rdfs:subClassOf :IndividualBias .
:ConsumerBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FConsumerBias"@en ;
rdfs:comment "Arises when an algorithm or platform provides users with a new venue within which to express their biases, and may occur from either side, or party, in a digital interaction.."@en ;
rdfs:label "consumer bias"@en ;
rdfs:subClassOf :IndividualBias .
:HumanReportingBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FHumanReportingBias"@en ;
rdfs:comment "When users rely on automation as a heuristic replacement for their own information seeking and processing."@en ;
rdfs:label "human reporting bias"@en ;
rdfs:subClassOf :IndividualBias .
:InterpretationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FInterpretationBias"@en ;
rdfs:comment "A form of information processing bias that can occur when users interpret algorithmic outputs according to their internalized biases and views."@en ;
rdfs:label "interpretation bias"@en ;
rdfs:subClassOf :IndividualBias .
:LossOfSituationalAwarenessBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FLossOfSituationalAwarenessBias"@en ;
rdfs:comment "When automation leads to humans being unaware of their situation such that, when control of a system is given back to them in a situation where humans and machines cooperate, they are unprepared to assume their duties. This can be a loss of awareness over what automation is and isn’t taking care of."@en ;
rdfs:label "loss of situational awareness bias"@en ;
rdfs:subClassOf :IndividualBias .
:ModeConfusionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FModeConfusionBias"@en ;
rdfs:comment "When modal interfaces confuse human operators, who misunderstand which mode the system is using, taking actions which are correct for a different mode but incorrect for their current situation. This is the cause of many deadly accidents, but also a source of confusion in everyday life."@en ;
rdfs:label "mode confusion bias"@en ;
rdfs:subClassOf :IndividualBias .
:PresentationBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FPresentationBias"@en ;
rdfs:comment "Biases arising from how information is presented on the Web, via a user interface, due to rating or ranking of output, or through users’ own self-selected, biased interaction."@en ;
rdfs:label "presentation bias"@en ;
rdfs:subClassOf :IndividualBias .
:RashomonEffectBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FRashomonEffectBias"@en ;
rdfs:comment "Refers to differences in perspective, memory and recall, interpretation, and reporting on the same event from multiple persons or witnesses."@en ;
rdfs:label "Rashomon effect bias"@en ;
rdfs:subClassOf :IndividualBias .
:SelectiveAdherenceBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FSelectiveAdherenceBias"@en ;
rdfs:comment "Decision-makers’ inclination to selectively adopt algorithmic advice when it matches their pre-existing beliefs and stereotypes."@en ;
rdfs:label "selective adherence bias"@en ;
rdfs:subClassOf :IndividualBias .
:StreetlightEffectBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FStreetlightEffectBias"@en ;
rdfs:comment "A bias whereby people tend to search only where it is easiest to look."@en ;
rdfs:label "streetlight effect bias"@en ;
rdfs:subClassOf :IndividualBias .
:UserInteractionBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FUserInteractionBias"@en ;
rdfs:comment "Arises when a user imposes their own self-selected biases and behavior during interaction with data, output, results, etc."@en ;
rdfs:label "user interaction bias"@en ;
rdfs:subClassOf :IndividualBias .
:InstitutionalBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FInstitutionalBias"@en ;
rdfs:comment "In contrast to biases exhibited at the level of individual persons, institutional bias refers to a tendency exhibited at the level of entire institutions, where practices or norms result in the favoring or disadvantaging of certain social groups. Common examples include institutional racism and institutional sexism."@en ;
rdfs:label "institutional bias"@en ;
rdfs:subClassOf :CognitiveBias .
:SunkCostFallacyBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FSunkCostFallacyBias"@en ;
rdfs:comment "A human tendency where people opt to continue with an endeavor or behavior due to previously spent or invested resources, such as money, time, and effort, regardless of whether costs outweigh benefits. For example, in AI, the sunk cost fallacy could lead development teams and organizations to feel that because they have already invested so much time and money into a particular AI application, they must pursue it to market rather than deciding to end the effort, even in the face of significant technical debt and/or ethical debt."@en ;
rdfs:label "sunk cost fallacy bias"@en ;
rdfs:subClassOf :GroupBias .
:SystemicBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FSystemicBias"@en ;
rdfs:comment "Biases that result from procedures and practices of particular institutions that operate in ways which result in certain social groups being advantaged or favored and others being disadvantaged or devalued."@en ;
rdfs:label "systemic bias"@en ;
rdfs:subClassOf :CognitiveBias .
:HistoricalBias rdf:type owl:Class ;
terms:source "https://bioportal.bioontology.org/ontologies/AIO/?p=classes&conceptid=https%3A%2F%2Fw3id.org%2Faio%2FHistoricalBias"@en ;