Towards Fairness in AI: A Systematic Mapping Study on Software Engineering Solutions
Bibliographic record
Abstract
Fairness in AI systems has become a crucial concern in software engineering, increasing attention since the field has evolved in the last few years. While some guidelines to address fairness have been proposed, achieving a comprehensive understanding of research solutions on fairness in AI models is trick. This paper presents a systematic literature mapping to explore and categorize the current advancements in fairness solutions in software engineering, focusing on two key elements: research trends and focus. We develop a classification framework to organize research on Software Fairness in a new perspective, applying it to 83 selected studies. As main result, we have found that researchers are creating practical AI solutions—specifically, framework solutions— focusing on group fairness metrics during the post-processing stage, these solutions are predominantly integrated during implementation and testing phases in software engineering process, with less attention to early stages like planning and design. Our results indicate a need to include fairness considerations throughout all stages of the software development lifecycle. Our analysis provides a comprehensive overview of the field, laying a foundation for guiding future research and practical applications of fairness in software systems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.052 | 0.187 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.032 | 0.022 |
| Science and technology studies | 0.004 | 0.006 |
| Scholarly communication | 0.007 | 0.012 |
| Open science | 0.002 | 0.006 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".