Skip to content

Commit 566cd9d

Browse files
Nkluge-correaNkluge-correa
committed
updated and added readme files
1 parent 4d10cd8 commit 566cd9d

File tree

14 files changed

+169
-15
lines changed

14 files changed

+169
-15
lines changed

‎AIA/README.md

+6-6
Original file line numberDiff line numberDiff line change
@@ -12,23 +12,23 @@ $$\text{Impact Score}\_{\text{ privacy}} = \frac{\text{Score}}{\text{Max Score}}
1212

1313
$$\text{AIA Score} = \frac{\text{Impact Score}\_{\text{ privacy}} + \text{Impact Score}\_{\text{ transparency}} + \text{Impact Score}\_{\text{ accountability}}}{\text{Nº of evaluated principles}}$$
1414

15-
In the following folders (`anti-discrimination`, `consumer-rights`, `privacy-and-data-protection`, and `protection-children`), you will find all questions for each AIA, presented as `.yaml` files, as shown in the figure above. These specification files are structured as shown below:
15+
In the following folders ([anti-discrimination](./anti-discrimination), [consumer-rights](./consumer-rights), [privacy-and-data-protection](./privacy-and-data-protection), and [protection-children](./protection-children)), you will find all questions for each AIA, presented as YAML files, as shown in the figure above. These specification files are structured as shown below:
1616

1717
```yaml
1818
question: "The question you wish to implement."
19-
multiple: false # if is not multiple choice, else, `true``
19+
multiple: false # If is not multiple choice, else, `true`.
2020
options:
2121
- Option 1
2222
- Option 2
2323
- Option 3
24-
scores: # the number of options and respective scores should match, and be in the same order
24+
scores: # The number of options and respective scores should match, and be in the same order.
2525
- 1.0
2626
- 2.0
2727
- 3.0
28-
principle1: true # if this question evaluates this principle, else, `false`
28+
principle1: true # If this question evaluates this principle, else, `false`.
2929
principle2: true
3030
principle3: true
31-
information: "An informative that may help the user in understanding the question." # if no information is to be given, set the value to `null`
31+
information: "An informative that may help the user understand the question." # If no information is to be given, set the value to `null`.
3232
```
3333
34-
You can create as many questions as you want. Only remember that this demo will display them in ascending order, using their number as a sorting key.
34+
You can create as many questions as you want. Remember that this demo will display them in ascending order, using their number as a sorting key.

‎AIA/anti-discrimination/README.md

+22
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,22 @@
1+
# AIA - Anti-discrimination & Trustworthiness
2+
3+
You will find all the questions for our impact assessment on **anti-discrimination and trustworthiness** here. All questions are stored as YAML files. These specification files are structured as shown below:
4+
5+
```yaml
6+
question: "The question you wish to implement."
7+
multiple: false # If is not multiple choice, else, `true`.
8+
options:
9+
- Option 1
10+
- Option 2
11+
- Option 3
12+
scores: # The number of options and respective scores should match, and be in the same order.
13+
- 1.0
14+
- 2.0
15+
- 3.0
16+
principle1: true # If this question evaluates this principle, else, `false`.
17+
principle2: true
18+
principle3: true
19+
information: "An informative that may help the user understand the question." # If no information is to be given, set the value to `null`.
20+
```
21+
22+
You can create as many questions as you want. Only remember that this demo will display them in ascending order, using their number as a sorting key.

‎AIA/consumer-rights/README.md

+22
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,22 @@
1+
# AIA - Consumer Rights
2+
3+
You will find all the questions for our impact assessment on **consumer rights** here. All questions are stored as YAML files. These specification files are structured as shown below:
4+
5+
```yaml
6+
question: "The question you wish to implement."
7+
multiple: false # If is not multiple choice, else, `true`.
8+
options:
9+
- Option 1
10+
- Option 2
11+
- Option 3
12+
scores: # The number of options and respective scores should match, and be in the same order.
13+
- 1.0
14+
- 2.0
15+
- 3.0
16+
principle1: true # If this question evaluates this principle, else, `false`.
17+
principle2: true
18+
principle3: true
19+
information: "An informative that may help the user understand the question." # If no information is to be given, set the value to `null`.
20+
```
21+
22+
You can create as many questions as you want. Only remember that this demo will display them in ascending order, using their number as a sorting key.

‎AIA/privacy-and-data-protection/README.md

+22
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,22 @@
1+
# AIA - Privacy & Data Protection
2+
3+
You will find all the questions for our impact assessment on **privacy and data protection** here. All questions are stored as YAML files. These specification files are structured as shown below:
4+
5+
```yaml
6+
question: "The question you wish to implement."
7+
multiple: false # If is not multiple choice, else, `true`.
8+
options:
9+
- Option 1
10+
- Option 2
11+
- Option 3
12+
scores: # The number of options and respective scores should match, and be in the same order.
13+
- 1.0
14+
- 2.0
15+
- 3.0
16+
principle1: true # If this question evaluates this principle, else, `false`.
17+
principle2: true
18+
principle3: true
19+
information: "An informative that may help the user understand the question." # If no information is to be given, set the value to `null`.
20+
```
21+
22+
You can create as many questions as you want. Only remember that this demo will display them in ascending order, using their number as a sorting key.

‎AIA/protection-children/README.md

+22
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,22 @@
1+
# AIA - Protection & Safety of Children and Adolescents
2+
3+
You will find all the questions for our impact assessment on the **protection and safety of children and adolescents** here. All questions are stored as YAML files. These specification files are structured as shown below:
4+
5+
```yaml
6+
question: "The question you wish to implement."
7+
multiple: false # If is not multiple choice, else, `true`.
8+
options:
9+
- Option 1
10+
- Option 2
11+
- Option 3
12+
scores: # The number of options and respective scores should match, and be in the same order.
13+
- 1.0
14+
- 2.0
15+
- 3.0
16+
principle1: true # If this question evaluates this principle, else, `false`.
17+
principle2: true
18+
principle3: true
19+
information: "An informative that may help the user understand the question." # If no information is to be given, set the value to `null`.
20+
```
21+
22+
You can create as many questions as you want. Only remember that this demo will display them in ascending order, using their number as a sorting key.

‎EPS/README.md

+2-2
Original file line numberDiff line numberDiff line change
@@ -1,12 +1,12 @@
11
# Ethical Framing and the WHY-SHOULD-HOW method
22

3-
This folder contains all principles available for ethical framing (_fairness, privacy, reliability, sustainability, transparency, and truthfulness_). After the evaluation stage, the EPS framework requires that human evaluators classify the system under consideration in an impact matrix. The matrix comprises three recommendation levels tailored to each impact level - **high**, **intermediate**, and **low**.
3+
This folder contains all principles available for ethical framing ([fairness](./fairness), [privacy](./privacy), [reliability](./reliability), [sustainability](./sustainability), [transparency](./transparency), [truthfulness](./truthfulness)). After the evaluation stage, the EPS framework requires that human evaluators classify the system under consideration in an impact matrix. The matrix comprises three recommendation levels tailored to each impact level - **high**, **intermediate**, and **low**.
44

55
<p align="center">
66
<img src="../img/eps-ethical-framing.png" alt="A graphical representation of an Algorithmic Impact Assessment and Ethical Framing in AI, focusing on human-centered design. The image consists of two main sections with pie charts and a grid layout, indicating levels of privacy, consumer rights, antidiscrimination, transparency, and fairness. A grey silhouette icon represents a human-centered approach. The legend explains the meaning of the colors." width="600px">
77
</p>
88

9-
Hence, each principle has three distinct possible recommendations tailored to specific impact levels, e.g., Sustainability-low, Sustainability-intermediate, and Sustainability-high.
9+
Hence, each principle has three distinct possible recommendations tailored to specific impact levels, e.g., [Sustainability-low](./sustainability/SHOULD-HOW-LOW.md), [Sustainability-intermediate](./sustainability/SHOULD-HOW-INTERMEDIATE.md), and [Sustainability-high](./sustainability/SHOULD-HOW-HIGH.md).
1010

1111
<p align="center">
1212
<img src="../img/eps-ethical-framing-2.png" alt="An infographic illustrating the criteria for Human-Centered Evaluation, with icons and text representing Fairness, Privacy, Reliability, Transparency, Truthfulness, and Sustainability. Each criterion is rated on a scale from low to high. The legend explains the meaning of the colors." width="600px">

‎EPS/fairness/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Fairness
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Fairness. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

‎EPS/privacy/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Privacy
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Privacy. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

‎EPS/reliability/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Reliability
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Reliability. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

‎EPS/sustainability/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Sustainability
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Sustainability. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

‎EPS/transparency/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Transparency
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Transparency. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

‎EPS/truthfulness/README.md

+11
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,11 @@
1+
# Truthfulness
2+
3+
This folder contains a set of tailored recommendations for improvements in AI Truthfulness. These recommendations come in a WHY-SHOULD-HOW format.
4+
5+
The WHY-SHOULD-HOW methodology is the format in which the evaluation outcome is presented. The WHY step is structured to demonstrate the relevancy of each principle, providing the conceptualization and highlighting paradigmatic cases of deficit implementation in a structure that answers the questions "_What is said principle?_" and "_Why should you care about it?_". The SHOULD and HOW are attached to streamline the normative guidance and the practical tools to address it.
6+
7+
<p align="center">
8+
<img src="../../img/eps-why-should-how.png" alt="A diagram illustrating the relationship between AI and privacy, with a focus on transparency and fairness levels. The image shows a progression from low to high levels of transparency, privacy, and fairness, leading to AI’s role in privacy. Questions about the importance of privacy and methods to protect it are highlighted." width="600px">
9+
</p>
10+
11+
The user can find a human-readable version of our recommendations as markdown files, while the HTML files are used to render our demo.

0 commit comments

Comments
 (0)