Customizable Tabular Access to Web Data Records for Convenient Low-vision Screen Magnifier Interaction
Hae-Na Lee, Vikas Ashok · 2022 · ACM Transactions on Accessible Computing · doi:10.1145/3517044
Summary
This paper introduces TableView+, a browser extension designed to address the significant challenges low-vision screen magnifier users face when browsing websites containing web data records—structured listings such as product catalogs, job postings, flight search results, and restaurant menus. Screen magnifiers, while essential for low-vision users, create a fundamental problem: they display only a small portion of the screen at high magnification levels (often 3x-16x), making it extremely difficult to compare information across multiple data records that span horizontally across the page. TableView+ automatically detects and extracts web data records from pages using the STEM (Suffix Tree-based Extraction Method) algorithm, then presents them in a compact, customizable tabular format. The system identifies repeating structural patterns in page DOM elements to locate data records and their constituent attributes. Users can then customize the table view by adjusting column widths, filtering out irrelevant attributes, sorting by any column, and changing font sizes. Critically, these customizations can be saved and automatically applied when revisiting the same website or similar sites, reducing the overhead of repeated configuration. The technical approach involves semantic similarity matching using Word2Vec embeddings to transfer saved customizations between similar websites—for example, applying Amazon product listing preferences to eBay listings. The extension works as a Chrome browser add-on that overlays the original webpage with the extracted tabular view.
Key findings
A user study with 25 low-vision participants (ages 23-73) with diverse conditions including glaucoma, diabetic retinopathy, optic atrophy, cataracts, and macular degeneration demonstrated substantial improvements. On unfamiliar websites, TableView+ reduced task completion time by 60.9% compared to using a screen magnifier alone. On familiar websites, the reduction was 60.04%. When TableView+ was preloaded with saved customizations, completion times improved by an additional 8.5% on unfamiliar sites. Usability metrics showed significant gains: System Usability Scale (SUS) scores increased by 25.3% compared to the screen magnifier baseline. NASA Task Load Index (TLX) scores decreased by 23.3%, indicating substantially lower cognitive load. Participants particularly valued the reduction in horizontal scrolling—a major pain point when comparing data records that span wide areas of the screen. Qualitative feedback highlighted that the compact tabular format eliminated the need to mentally track information while scrolling back and forth, and the customization persistence meant users could quickly access preferred views on frequently visited sites.
Relevance
This research addresses a gap in assistive technology for low-vision users that existing tools fail to solve. While screen readers linearize content for blind users and screen magnifiers enlarge content, neither adequately supports the comparison tasks essential to everyday web activities like shopping, job searching, or booking travel. The 60%+ improvement in task completion time represents a meaningful quality-of-life enhancement. For practitioners, this work demonstrates the value of task-specific accessibility tools rather than one-size-fits-all solutions. The customization persistence feature shows how reducing setup overhead increases adoption and effectiveness. Organizations with customer-facing websites containing data records should consider how their layouts affect screen magnifier users and whether structured data markup could enable similar extraction tools. The research also validates that low-vision accessibility requires distinct solutions from blind accessibility—a reminder that visual impairment exists on a spectrum requiring diverse approaches.
Tags: low vision · screen magnifier · web accessibility · data extraction · browser extension · tabular data · comparison tasks · space compaction