3.6 KiB
3.6 KiB
type:: REVIEWS tags:: year:: 2025 venue:: TOSEM full-title:: Addressing Visual Impairments with Model-Driven Engineering: A Systematic Literature Review date-start:: 18-11-2025 - 07:07 date-submitted:: external-links:: status:: DONE deadline-submission:: file:: @TOSEM-2025-0885 parent:: todoist:: https://app.todoist.com/app/task/tosem-2025-0885-reviewer-agreed-6f4vvMGPcFvWHVJg
- SUMMARY
- The paper presents a systematic literature review on the application of Model-Driven Engineering to address accessibility needs of visually impaired users. From an initial set of 447 papers, 30 primary studies were selected and analyzed with respect to different aspects including considered impairments, employed modeling techniques, adopted evaluation practices, and reported strengths and limitations. The review highlights that most studies operate at a high level of abstraction, rely extensively on WCAG guidelines, provide limited implementation details, and seldom involve visually impaired end users.
- COMMENTS
- The paper is about an important and interesting topic discussing the outcomes of a systematic review effort. However, my main concerns about the paper are related to the following points:
- A first overall impression is that the manuscript is verbose. The clarity of the contribution would improve significantly by reducing descriptive passages and sharpening the analytical core of the review. For instance, the decreasing trend in publications and the low impact within MDE venues are noted, but the paper does not explain why these observations matter. Is the decline due to intrinsic difficulty? A shift in the community? A mismatch between MDE and accessibility needs?
- After reading the paper, I'm still questioning what an effective MDE contribution for visual impairments should entail. The authors note that most works overlook the real nature of visual impairments and provide only abstract role models. However, the paper could be more explicit about what it expected to see from an MDE perspective: how should modeling languages encode impairment-specific requirements? What does "accessibility challenges" mean in practice? What would a meaningful MDE-based pipeline look like? Without articulating this ideal perspective, the criticism remains vague.
- Table 1 introduces a scoring system apparently defined by the authors, but the paper does not explain how the criteria were derived, whether existing SLR guidelines influenced them, or whether a structured elicitation process was followed. A short methodological paragraph would strengthen transparency.
- Authors repeatedly claim that approaches lack reusability, reusable components, or reusable transformations. However, given the diversity of impairments, contexts, user needs, application domains, and UI constraints, it is not obvious what reusability should realistically mean in this space. The paper should clarify what kinds of artifacts could be reused (e.g., domain-independent modeling constructs, modular transformation patterns, template structures) and what cannot. Without this distinction, the critique risks becoming generic and not fully justified.
- The authors observe that DSLs generally capture UI structure rather than accessibility; this is a valuable point, but the analysis should identify what specific modeling constructs are missing and what an accessibility-aware DSL would need to express.
- Some structural issues also affect readability. Important figures are referenced long before they appear (e.g., Figure 4), breaking the flow. Certain sentences are abstract or vague ("robust", "transparent", "technically grounded work") and need clarification.