The accessibility and readability of Generative Artificial Intelligence systems like GPT and Google BARD are crucial factors that require thorough examination. In today’s digitally connected world, where AI-generated content is becoming increasingly prevalent across various sectors, ensuring that such content is accessible to a broad audience, easy to understand, and user-friendly is essential. This research was motivated by the growing influence of AI-generated content on various user groups, including individuals with varying levels of familiarity with technology and diverse cognitive abilities. This paper comprehensively assesses and draws comparisons between the accessibility and readability of Google BARD and GPT to identify areas for enhancement. GPT exhibits various accessibility issues across different WCAG 2.0 conformance levels, contrasting with Google BARD, which shows no accessibility problems. The results underscore the importance of addressing accessibility concerns in AI models to ensure inclusivity and user-friendliness for individuals with disabilities. Google BARD consistently provides more readable content than GPT, underscoring the importance of selecting the appropriate tool based on the target audience and desired readability.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Comparative Assessment of Accessibility and Readability in Generative AI: A Case Study of GPT and Google BARD

  • C. P. Afsal,
  • K. S. Kuppusamy

摘要

The accessibility and readability of Generative Artificial Intelligence systems like GPT and Google BARD are crucial factors that require thorough examination. In today’s digitally connected world, where AI-generated content is becoming increasingly prevalent across various sectors, ensuring that such content is accessible to a broad audience, easy to understand, and user-friendly is essential. This research was motivated by the growing influence of AI-generated content on various user groups, including individuals with varying levels of familiarity with technology and diverse cognitive abilities. This paper comprehensively assesses and draws comparisons between the accessibility and readability of Google BARD and GPT to identify areas for enhancement. GPT exhibits various accessibility issues across different WCAG 2.0 conformance levels, contrasting with Google BARD, which shows no accessibility problems. The results underscore the importance of addressing accessibility concerns in AI models to ensure inclusivity and user-friendliness for individuals with disabilities. Google BARD consistently provides more readable content than GPT, underscoring the importance of selecting the appropriate tool based on the target audience and desired readability.