Understanding Research Themes and Interactions at Scale within Blind and Low-vision Research in ACM and IEEE
Maximiliano Jeanneret Medina, Yong-Joon Thoo, Cédric Baudet, Jon E. Froehlich, Nicolas Ruffieux, Denis Lalanne · 2025 · ACM Transactions on Accessible Computing · doi:10.1145/3726531
Summary
This 46-page extended article builds on a 2023 ASSETS paper to provide a comprehensive, field-agnostic analysis of blind and low-vision (BLV) research published in ACM and IEEE venues between 2010 and 2022. Using a mixed-methods approach combining quantitative bibliometrics with qualitative analysis, the authors examined 880 highly-cited papers to map the landscape of BLV computing research. The methodology employed Document Bibliographic Coupling Analysis (DBCA) to cluster papers by shared references, combined with Title/Abstract/Keywords (TAK) analysis using natural language processing. The 100 most-cited papers received in-depth qualitative coding across multiple dimensions: research context, communities of focus, technological trends, and interaction modalities. The authors also performed a comparative analysis of 12 prior systematic literature reviews in the BLV field, published between 2010 and 2022. The study identifies four major research areas: Accessibility at Home & on the Go (N=280), covering access to digital media, activities of daily living, and visual assistance technologies like VizWiz and Be My Eyes; Non-Visual Interaction (N=195), focusing on touch, auditory, and multimodal interfaces; Education (N=54), emphasizing collaborative learning and STEM accessibility; and Orientation & Mobility (N=331), the largest area, addressing navigation assistance, obstacle avoidance, and mobility training through technologies like wearables and computer vision.
Key findings
The research reveals significant technological trends: mobile technologies dominate (N=224), followed by computer vision (N=116), wearables (N=106), navigation assistance (N=106), AI (N=70), and web applications (N=67). Mixed reality and AI applications have grown notably since 2015-2016. The top five devices are mobile devices (N=33), wearables (N=25), personal computers (N=17), VR headsets (N=5), and tangible objects (N=5). A critical finding concerns how "interaction" is conceptualized in BLV research. Analyzing 253 sentences containing the term across 880 abstracts, the authors found papers predominantly focus on Style modifiers (N=148)—describing technologies, modalities, and actions—rather than Quality modifiers (N=36) addressing how interactions feel or their effectiveness. This suggests the field emphasizes technical implementation over experiential qualities like usability, accessibility, and user satisfaction. The study highlights terminology inconsistencies: papers use varied terms ("blind," "low-vision," "visually impaired") without clear definitions. Research on language preferences shows 48.6% of disabled people prefer identity-first language versus 33% for person-first. Of 12 prior SLRs analyzed, only 2 specifically focused on low-vision individuals, revealing a significant research gap. Output modalities favor non-speech audio (N=46), haptics (N=18), and speech (N=10), while input modalities favor touch (N=32) and speech (N=11). Visual substitution strategies (N=61) far outnumber enhancement strategies (N=8).
Relevance
This paper serves as an essential roadmap for researchers and practitioners entering or working in BLV accessibility. The four-area framework (Accessibility at Home & on the Go, Non-Visual Interaction, Education, Orientation & Mobility) provides a structured way to understand where different technologies and research questions fit within the broader field. For practitioners, the finding that research emphasizes interaction Style over Quality suggests opportunities to focus more on user experience metrics beyond task completion. The overwhelming preference for visual substitution over enhancement strategies (61 vs. 8 papers) may overlook the needs and desires of low-vision users who retain residual vision. The comparative analysis of 12 prior SLRs identifies gaps—particularly in low-vision-specific research, longitudinal studies, and geographical analyses. The paper raises important methodological considerations: most prior SLRs focused on specific technologies or domains rather than providing field-wide perspectives. Future research trajectories identified include exploring novel AI-powered technologies (GPT-4, multimodal LLMs), developing ability-based design methodologies, conducting umbrella reviews synthesizing existing SLRs, and examining how BLV research has evolved over time across different geographic regions and research communities.
Tags: systematic literature review · blind and low vision · bibliometrics · mixed methods · research themes · interaction design · visual impairment · assistive technology