An Accessibility Evaluation Platform: Borrowing from Web 2.0
Yui-Liang Chen, Gina Lin · 2008 · Proceedings of the 2008 International Cross-Disciplinary Conference on Web Accessibility (W4A) · doi:10.1145/1368044.1368067
Summary
This paper proposes a community-driven Accessibility Evaluation Platform for Taiwan, drawing on Web 2.0 principles to supplement the government-led accessibility compliance program administered by the Research, Development & Evaluation Commission (RDEC). At the time of writing, Taiwan's Web Accessibility Regulation (established 2003) had led to 4,325 websites applying for accessibility conformance by August 2007, representing about 73% of public sector sites. However, the RDEC was the sole body conducting accessibility reviews, creating a bottleneck — progress reports from 2006 and 2007 showed little improvement between years. The authors argue that the RDEC's centralized approach could not scale, and that crowdsourcing evaluations through a Web 2.0-style platform would accelerate progress. The conceptual framework rests on two hypotheses drawn from social learning theory (Vygotsky) and social networking theory (Milgram): that increased interaction improves learning outcomes, and that social networks attract users who would otherwise not engage with accessibility. The proposed platform architecture consists of front-end and back-end interfaces. The front-end allows content developers to register websites for conformance review, maintain blogs about accessible design, and receive community feedback. Community members can conduct human reviews, publish ePapers, and participate in forums. The back-end provides administrators with member identity management, ePaper publishing, blog management, integration feedback, and forum management systems.
Key findings
The paper's primary contribution is conceptual rather than empirical — no implementation or evaluation of the platform is reported. The core insight is that Taiwan's centralized, government-only model for accessibility evaluation was not scaling: annual progress reports showed stagnation between 2006 and 2007 despite the existence of the Freego automated validation tool and RDEC support services. The proposed platform incorporates a member ranking system where reviewers accumulate credibility points through accurate evaluations, with peer review mechanisms that downgrade rankings for incorrect assessments. This reputation system was designed to maintain quality while opening participation. The platform would generate "Top 10" and "Bottom 10" website lists distributed via ePaper to subscribers and content developers, creating public accountability. The authors ground their proposal in collaborative learning research showing that group interaction produces higher performance and productivity than individual work.
Relevance
While this paper is primarily a conceptual proposal from 2008 without implementation data, it anticipates ideas that have since become mainstream in accessibility practice — particularly community-driven evaluation, crowdsourced auditing, and the limitations of purely automated or centralized compliance checking. The Taiwan context illustrates a challenge that remains universal: government accessibility mandates outpace the capacity to evaluate and enforce them. The idea of combining automated tools (like Taiwan's Freego) with structured community human review prefigures modern approaches that blend automated scanning with expert manual testing. The reputation-based quality control mechanism for community reviewers addresses a real concern about crowdsourced accessibility evaluation — ensuring review quality without restricting participation. For organizations scaling their accessibility programs, the paper's core argument remains valid: no single team can evaluate all digital properties, and structured community involvement is essential for comprehensive coverage.
Tags: accessibility evaluation · web 2.0 · crowdsourcing · community review · Taiwan · government accessibility · social learning · accessibility policy
Standards referenced: WCAG