This chapter proposes a novel approach to the development and assessment of English disaster and environmental risk information for people from diverse language and cultural backgrounds who require readable, more accessible translations, regardless of education level, and cultural or linguistic background.¬¬¬ To illustrate the development of machine learning classifiers for the purpose of the predictive assessment of the likelihood that an English text will be translated into an accessible language, we will use Japanese as an illustrative case study, given the language and cultural contrast between English and Japanese.