Case ReportPublished systematic mapping study2025
AAB-CASE-2025-RV-057
Fairness for machine learning software in education: A systematic mapping study
Journal of Systems & Software; Vietnam / Norway team.
This page documents an AI literacy or AI education case for registry purposes. It is descriptive and does not imply AAB endorsement of any specific tool, provider, or intervention.
01
Implementation
Universities
02
Learning context
Private program
03
AI role
Evaluator
04
Outcome signal
Equity
Registry Facets
0
Education Level
- Higher education
Subject Area
- Learning analytics
- Ethics and society
Use Case Type
- Systematic review
Stakeholder Group
- Researchers
AI Capability Type
- Prediction / risk scoring
Implementation Model
- System-level guidance
Evidence Type
- Systematic review
Outcomes Domain
- Equity
- Policy
Implementing Organization
1
Organization Type
Universities
Location
Viet Nam / Norway
Primary Facilitator Role
Authors
Learning Context
2
Setting Type
- Private program
Session Format
Systematic mapping of literature
Duration
2002–2023 corpus
Group Size
63 studies
Devices
Educational ML systems
Constraints
- Publication bias
- Evolving ML paradigms
Learner Profile
3
Age Range
Higher education populations in reviewed work
Prior AI Exposure Assumed
N/A
Prior Programming Background Assumed
N/A
Educational Intent
4
Primary Learning Goals
- Map ML fairness research in education software
Secondary Learning Goals
- Identify gaps: multi-sensitive attributes, metric combinations, benchmarks
What This Was Not
- Not empirical fairness audit of one product
AI Tool Description
5
Tool Type
ML fairness in educational deployments (review)
AI Role
- Evaluator
- Automation tool
Languages
Global literature
User Interaction Model
Safeguards
- Fairness testing pipelines for production EdML
Activity Design
6
Activity Flow
- Define RQs
- Screen corpus
- Synthesize themes
Human Vs AI Responsibilities
Scaffolding Strategies
Observed Challenges
7
Educators Reported
- Dominance of classical tabular ML vs modern deep models
- Need open benchmarks
Design Adaptations
8
Adaptations
- Consolidates definitions and sensitive variables used in field
Reported Outcomes
9
Engagement
Learning Signals
Educators Reflection
Roadmap for engineers and IRBs on educational ML fairness.
Ethical & Privacy Considerations
10
Privacy
- Protected attributes in training data
- Discrimination risk
Evidence Type
11
Evidence
- Activity documentation
- Practitioner observation
Relevance to Research
12
Potential Research Use
- Fairness assurance under intersectionality
- Open-source EdML fairness benchmarks
Relevant Research Domains
- Algorithmic fairness
- Educational data mining
Case Status
13
Case Status
- Completed
AAB Classification Tags
14
Age
HE students (reviewed)
Setting
Global HE
AI Function
Fair ML in education
Pedagogy
Mapping study
Risk Level
High if unfair models deployed
Data Sensitivity
High
Registry Metadata
15
Case ID
AAB-CASE-2025-RV-057
Publication Status
Published systematic mapping study
Tags
caseHigher educationViet Nam / NorwaySystem-level guidancePrediction / risk scoringLearning analyticsEthics and societySystematic review