Skip to content

Commit 4bfe37c

Browse files
jfpariejfparie
committed
added new VD photo and new AI expertises
1 parent 1734b80 commit 4bfe37c

File tree

12 files changed

+271
-479
lines changed

12 files changed

+271
-479
lines changed

‎content/english/_index.md‎

Lines changed: 104 additions & 129 deletions
Original file line numberDiff line numberDiff line change
@@ -1,88 +1,4 @@
11
---
2-
Areas_of_AI_expertise:
3-
title: AI expertise
4-
enable: true
5-
width_m: 4
6-
width_s: 12
7-
button_text: Discuss collaboration
8-
button_link: /knowledge-platform/collaboration/
9-
feature_item:
10-
- name: Profiling
11-
icon: fas fa-chart-pie
12-
content: >
13-
Auditing rule-based and ML-driven profiling, e.g., differentiation
14-
policies, selection criteria, Z-testing, model validation and
15-
organisational aspects
16-
- name: FP-FN balancing
17-
icon: fa-solid fa-scale-balanced
18-
content: >
19-
Context-dependent review of ML and DL confusion matrix-based evaluation
20-
metrics, such as False Positives (FPs) and False Negatives (FNs)
21-
- name: Ranking
22-
icon: fas fa-ranking-star
23-
content: >
24-
Recommender systems are everywhere. With the new Digital Services Act
25-
(DSA) came into forces last summer, auditing ranking systems is highly
26-
relevant
27-
Building_algoprudence:
28-
title: Building _algoprudence_
29-
button_text: Our working method
30-
button_link: /algoprudence/how-we-work/
31-
steps:
32-
- title: Indentifying issue
33-
content: >
34-
Identifying a concrete ethical issue in a real algorithm or
35-
data-analysis tool
36-
- title: Problem statement
37-
content: >
38-
Describe ethical issue, legal aspects and hear stakeholders and affected
39-
groups
40-
- title: Advice commission
41-
content: >
42-
Deliberative conversation on ethical issue by diverse and inclusive
43-
advice commission
44-
- title: Public advice
45-
content: >
46-
Advice of commission is published together with problem statement on our
47-
website. Publicly sharing the problem statement and normative advice is
48-
called *algoprudence*
49-
Advantages_of_algoprudence:
50-
title: Advantages of algoprudence
51-
image: /images/other/Awb.jpg
52-
button_text: White paper on algoprudence
53-
feature_item:
54-
- name: Learn & harmonize
55-
icon: fas fa-book-reader
56-
content: >
57-
[<span style="color:#005aa7">></span>](/algoprudence) Ignite collective
58-
learning process to deploy and audit responsible AI
59-
60-
61-
[<span style="color:#005aa7">></span> ](/algoprudence)Harmonizes the
62-
resolution of ethical questions and the interpretation of open legal
63-
norms
64-
- name: Question & criticize
65-
icon: fas fa-comment
66-
content: >
67-
[<span style="color:#005aa7">></span>](/algoprudence) Fostering
68-
criticism on normative decision-making through transparency
69-
70-
71-
[<span style="color:#005aa7">></span> ](/algoprudence)Informing public
72-
debate with important ethical issues to be discussed within democratic
73-
sight
74-
- name: Inclusion & participation
75-
icon: fas fa-hands-helping
76-
content: >
77-
[<span style="color:#005aa7">></span> ](/algoprudence)Connecting various
78-
stakeholders to design ethical algorithms together with technical
79-
experts
80-
81-
82-
[<span style="color:#005aa7">></span> ](/algoprudence)European answer to
83-
deploy responsible AI systems
84-
Title_gif:
85-
title: Jurisprudence for algorithms
862
Banner:
873
image: /images/svg-illustrations/main_illustration.svg
884
tag: definition
@@ -185,6 +101,52 @@ Distinctive_in:
185101
style="color:white;text-decoration: underline;">publicly available</a>,
186102
increasing collective knowledge how to deploy and use algorithms in an
187103
responsible way
104+
Areas_of_AI_expertise:
105+
title: AI expertise
106+
enable: true
107+
width_m: 4
108+
width_s: 12
109+
button_text: Discuss collaboration
110+
button_link: /knowledge-platform/collaboration/
111+
feature_item:
112+
- name: Algorithms for decision support
113+
icon: fas fa-divide
114+
content: >
115+
Audting data-analysis methods and algorithms used for decision support.
116+
Among others by checking organizational checks and balances, and
117+
assessing the quantitative dimension
118+
- name: AI Act standards
119+
icon: fas fa-certificate
120+
content: >
121+
As Algorithm Audit is part of Dutch and Europen standardization
122+
organisation NEN and CEN-CENELEC, AI systems are audited according to
123+
the latest standards. See also our public <a
124+
href="https://algorithmaudit.eu/knowledge-platform/standards/"
125+
style="text-decoration: underline;">knowledge base</a> on
126+
standardization
127+
- name: Profiling
128+
icon: fas fa-chart-pie
129+
content: >
130+
Auditing rule-based and ML-driven profiling, e.g., differentiation
131+
policies, selection criteria, Z-testing, model validation and
132+
organisational aspects
133+
- name: FP-FN balancing
134+
icon: fa-solid fa-scale-balanced
135+
content: >
136+
Context-dependent review of ML and DL confusion matrix-based evaluation
137+
metrics, such as False Positives (FPs) and False Negatives (FNs)
138+
- name: Ranking
139+
icon: fas fa-ranking-star
140+
content: >
141+
Recommender systems are everywhere. With the new Digital Services Act
142+
(DSA) came into forces last summer, auditing ranking systems is highly
143+
relevant
144+
- name: Generative AI
145+
icon: fas fa-robot
146+
content: >
147+
Auditing training process of foundation model, among others selection of
148+
training data, human feedback for reinformcement learning and risk
149+
management, according AI Act standards standaarden
188150
Recent_audits:
189151
title: Recent audits
190152
feature_item:
@@ -202,55 +164,68 @@ Recent_audits:
202164
content: >
203165
Assessment of risk distributions through Z-tests and bias test for
204166
various steps in algorithmic-driven decision-making process
205-
Title_video:
206-
title: The Movie
207-
video_mp4: /videos/AA_video_(1080p).mp4
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-
167+
Building_algoprudence:
168+
title: Building _algoprudence_
169+
button_text: Our working method
170+
button_link: /algoprudence/how-we-work/
171+
steps:
172+
- title: Indentifying issue
173+
content: >
174+
Identifying a concrete ethical issue in a real algorithm or
175+
data-analysis tool
176+
- title: Problem statement
177+
content: >
178+
Describe ethical issue, legal aspects and hear stakeholders and affected
179+
groups
180+
- title: Advice commission
181+
content: >
182+
Deliberative conversation on ethical issue by diverse and inclusive
183+
advice commission
184+
- title: Public advice
185+
content: >
186+
Advice of commission is published together with problem statement on our
187+
website. Publicly sharing the problem statement and normative advice is
188+
called *algoprudence*
189+
Advantages_of_algoprudence:
190+
title: Advantages of algoprudence
191+
image: /images/knowledge_base/NJB-cover.jpg
192+
button_text: White paper on algoprudence
193+
button_link: /knowledge-platform/knowledge-base/njb-artikel/
194+
feature_item:
195+
- name: Learn & harmonize
196+
icon: fas fa-book-reader
197+
content: >
198+
[<span style="color:#005aa7">></span>](/algoprudence) Ignite collective
199+
learning process to deploy and audit responsible AI
250200
251201
202+
[<span style="color:#005aa7">></span> ](/algoprudence)Harmonizes the
203+
resolution of ethical questions and the interpretation of open legal
204+
norms
205+
- name: Question & criticize
206+
icon: fas fa-comment
207+
content: >
208+
[<span style="color:#005aa7">></span>](/algoprudence) Fostering
209+
criticism on normative decision-making through transparency
252210
253211
212+
[<span style="color:#005aa7">></span> ](/algoprudence)Informing public
213+
debate with important ethical issues to be discussed within democratic
214+
sight
215+
- name: Inclusion & participation
216+
icon: fas fa-hands-helping
217+
content: >
218+
[<span style="color:#005aa7">></span> ](/algoprudence)Connecting various
219+
stakeholders to design ethical algorithms together with technical
220+
experts
254221
255222
223+
[<span style="color:#005aa7">></span> ](/algoprudence)European answer to
224+
deploy responsible AI systems
225+
Title_gif:
226+
title: Jurisprudence for algorithms
227+
Title_video:
228+
title: The Movie
229+
video_mp4: /videos/AA_video_(1080p).mp4
230+
---
256231

‎content/english/algoprudence/how-we-work.md‎

Lines changed: 1 addition & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -1,22 +1,16 @@
11
---
2-
background_color: '#ffffff'
32
title: How we work
43
subtitle: >
54
Over the years, we have developed our own deliberative audit methodology. We
65
elaborate below on our 4-step working method.
76
image: /images/svg-illustrations/main_illustration.svg
8-
text_field1:
9-
title: Our 4-step audit methodology
10-
icon: fas fa-eye
11-
id: methodology
12-
content: ''
137
team:
148
title: Algoprudence team
159
icon: fas fa-user-friends
1610
button_text: Other teams
1711
button_link: /about/teams/
1812
team_members:
19-
- image: /images/people/VDjwalapersad.png
13+
- image: /images/people/VDjwalapersad.jpeg
2014
name: Vardâyani Djwalapersad
2115
bio: |
2216
Project manager Algoprudence

‎content/english/algoprudence/submit-a-case.md‎

Lines changed: 43 additions & 51 deletions
Original file line numberDiff line numberDiff line change
@@ -6,53 +6,6 @@ subtitle: >
66
algorithm. Or submit a carefully documented normative judgement for inlcusion
77
in our case repository.
88
image: /images/svg-illustrations/case_repository.svg
9-
team:
10-
title: Team Algoprudence
11-
icon: fas fa-user-friends
12-
button_text: Other teams
13-
button_link: /about/teams/
14-
team_members:
15-
- image: /images/people/VDjwalapersad.png
16-
name: Vardâyani Djwalapersad
17-
bio: |
18-
Project manager Algoprudence
19-
- image: /images/people/JParie.jpg
20-
name: Jurriaan Parie
21-
bio: |
22-
Director-board member
23-
- image: /images/people/AVoogt.jpeg
24-
name: Ariën Voogt
25-
bio: |
26-
Board member
27-
- image: /images/people/SMohammad.jpeg
28-
name: Samaa Mohammad-Ulenberg
29-
bio: |
30-
Board member
31-
reports_preview:
32-
title: Recent audits
33-
icon: fas fa-file
34-
button_text: Case repository
35-
button_link: /algoprudence
36-
feature_item:
37-
- name: 'Risk Profiling Social Welfare Re-examination '
38-
image: /images/algoprudence/AA202302/AA202302A_cover_EN.png
39-
link: >-
40-
/algoprudence/cases/aa202302_risk-profiling-for-social-welfare-reexamination/
41-
content: >
42-
Normative advice commission provides rationales why these variables are
43-
eligible or not as a profiling selection criterion for a xgboost
44-
algorithm
45-
- name: Technical audit indirect discrimination
46-
image: /images/algoprudence/AA202401/Cover.png
47-
link: /algoprudence/cases/aa202401_bias-prevented/
48-
content: >
49-
Qualitative and quantitative audit of algorithmic-driven college
50-
allowances control process. Assessment of risk distributions through
51-
Z-tests and bias test in the control process
52-
text_field1:
53-
title: What are the next steps after case submission?
54-
icon: fas fa-forward
55-
content: ''
569
form1:
5710
title: Case information
5811
content: ''
@@ -142,12 +95,51 @@ form2:
14295
described above, kept for the minimum necessary duration, and is
14396
securely stored in a protected environment</span>
14497
id: terms-conditions
145-
values:
146-
- label: Agree
147-
value: agree
148-
id: agree
14998
required: true
15099
type: checkbox
100+
reports_preview:
101+
title: Recent audits
102+
icon: fas fa-file
103+
button_text: Case repository
104+
button_link: /algoprudence
105+
feature_item:
106+
- name: 'Risk Profiling Social Welfare Re-examination '
107+
image: /images/algoprudence/AA202302/AA202302A_cover_EN.png
108+
link: >-
109+
/algoprudence/cases/aa202302_risk-profiling-for-social-welfare-reexamination/
110+
content: >
111+
Normative advice commission provides rationales why these variables are
112+
eligible or not as a profiling selection criterion for a xgboost
113+
algorithm
114+
- name: Technical audit indirect discrimination
115+
image: /images/algoprudence/AA202401/Cover.png
116+
link: /algoprudence/cases/aa202401_bias-prevented/
117+
content: >
118+
Qualitative and quantitative audit of algorithmic-driven college
119+
allowances control process. Assessment of risk distributions through
120+
Z-tests and bias test in the control process
121+
team:
122+
title: Team Algoprudence
123+
icon: fas fa-user-friends
124+
button_text: Other teams
125+
button_link: /about/teams/
126+
team_members:
127+
- image: /images/people/VDjwalapersad.jpeg
128+
name: Vardâyani Djwalapersad
129+
bio: |
130+
Project manager Algoprudence
131+
- image: /images/people/JParie.jpg
132+
name: Jurriaan Parie
133+
bio: |
134+
Director-board member
135+
- image: /images/people/AVoogt.jpeg
136+
name: Ariën Voogt
137+
bio: |
138+
Board member
139+
- image: /images/people/SMohammad.jpeg
140+
name: Samaa Mohammad-Ulenberg
141+
bio: |
142+
Board member
151143
---
152144

153145
{{< tab_header width="6" tab1_id="case-for-review" default_tab="case-for-review" tab1_title="Submit a case for normative review" tab2_id="case-repository" tab2_title="Add a case to algoprudence repository" >}}

0 commit comments

Comments
 (0)