Self-regulated and Participatory Automatic Text Simplification

This contribution draws upon conceptual and theoretical insights that have emerged from the authors’ research project (Open-LS) aiming at the development of an application-oriented framework for comprehensibility enhanced text formats tailored to individuals with low literacy level mainly due to cognitive impairments. We address general and domain-specific regulatory model requirements for text-simplifying Neural Language Models for use in assistive applications in German. Against this background, we position ourselves as advocates of an accessibility-by-hypertext approach. Further, we put forward for discussion building blocks for an AI Act-compliant self-regulated framework for participation-driven text simplification systems.