MokE: A Tool for Mobile-ok Evaluation of Web Content
John Garofalakis, Vassilios Stefanis · 2008 · Proceedings of the 2008 International Cross-Disciplinary Conference on Web Accessibility (W4A) · doi:10.1145/1368044.1368058
Summary
This paper presents MokE (Mobile-OK Evaluator), a web-based tool for evaluating whether web content provides an acceptable experience on mobile devices, based on W3C's mobileOK Basic Tests and Mobile Web Best Practices. Written at a time when mobile web access was rapidly growing — particularly in developing countries where mobile users already outnumbered desktop users — MokE addressed the challenge that most web content was authored for desktop browsers without considering mobile constraints such as small screens (minimum 120 pixels), limited bandwidth (20KB page weight limit), no scripting support, and restricted input mechanisms. The tool consists of four modules: a web crawler that traverses a site and collects pages, a hidden web module that discovers form-based dynamic content not reachable by following links, a mobileOK checker module built on the W3C's Java-based mobileOK checker library that runs 26 automated tests, and a rating module that calculates overall site scores. The W3C's Mobile Web Best Practices contain 60 recommendations organized across five categories: Overall Behavior, Navigation and Links, Page Layout and Content, Page Definition, and User Input. The mobileOK Basic Tests are a machine-verifiable subset of these practices, testing for issues such as auto-refresh, proper caching headers, UTF-8 encoding, valid markup, image sizing, use of tables for layout, pop-ups, page size limits, and text alternatives for non-text content.
Key findings
MokE's distinctive contribution was its hidden web module, which addressed a significant gap in existing mobile evaluation tools. The "hidden web" (or deep web) — content behind forms that is dynamically generated from databases — was estimated at several million pages and growing rapidly. Standard crawlers cannot reach this content because it requires form submission. MokE's approach identifies forms in crawled pages, allows users to provide query terms, and then submits those forms to discover and test the dynamically generated content. The tool calculates two types of ratings: a basic rating R on a 0-10 scale that aggregates pass/fail/warn results across all pages and tests, and a weighted rating WR that allows users to assign importance weights (1-5 Likert scale) to each of the 26 tests, creating customizable evaluation profiles. In the example evaluation shown, 8 crawled pages produced 208 test results (8 pages x 26 tests), with 87.5% passing, 12.02% failing, and 0.48% returning warnings. The tool was built on Apache Web Server with PHP 5 and Java 1.5, using MySQL for storage, and the W3C's open-source mobileOK checker library as its testing engine.
Relevance
While the specific mobile constraints described in this 2008 paper have been largely superseded by modern smartphones with full browsers, the underlying principles remain relevant to mobile accessibility today. The W3C's "one web" philosophy — that content should be accessible regardless of device — anticipated the responsive design movement and continues to underpin mobile accessibility requirements in WCAG 2.1 and 2.2 (which added mobile-specific success criteria). The paper's focus on the hidden web highlights a challenge that persists in modern accessibility testing: dynamically generated content from single-page applications, API-driven interfaces, and database-backed pages is difficult for automated tools to discover and evaluate. MokE's approach of combining automated crawling with user-guided form interaction to reach dynamic content foreshadows contemporary approaches to testing JavaScript-heavy web applications. The customizable weighting system for evaluation profiles also anticipates the need for context-specific accessibility evaluation rather than one-size-fits-all compliance checking.
Tags: mobile accessibility · mobile web · automated testing · web crawling · W3C standards · evaluation tools · hidden web
Standards referenced: W3C Mobile Web Best Practices · W3C mobileOK Basic Tests