Klartext | Michael Färber's Research Group

Klartext builds an AI-assisted workflow to translate web content into plain / easy language (Leichte/Einfache Sprache).
The focus is on making the websites of ScaDS.AI and TU Dresden more accessible, without turning the process into a manual, page-by-page rewrite.

Updates

Our demo paper related to Klartext has been published and presented at ECIR 2025:
SimplifyMyText: An LLM-Based System for Inclusive Plain Language Text Simplification (Färber et al., 2025)

What we do

Klartext combines modern language models with a practical publishing workflow:

We extract and normalize website content into a clean text representation that can be translated reliably at scale.
We generate plain/easy-language versions with LLMs and keep a human-in-the-loop step for confirmation and quality checks before publication.
We build an intuitive, lightweight tool (open source) that supports continuous improvements through user feedback and iterative refinement.

The project is designed as a pilot with a clear path towards broader adoption across further TU Dresden web content.

Why it matters

A significant number of people struggle with complex written language, including individuals with reading difficulties, cognitive impairments, or limited proficiency in German. At the same time, public and scientific institutions increasingly communicate through websites — and these sites are often not accessible to everyone.

Plain/easy-language versions reduce barriers and enable broader participation in scientific and cultural discourse. However, producing and maintaining such content manually is expensive and does not scale well. Klartext therefore targets the practical gap: high-quality plain-language content, created efficiently and maintained sustainably with an AI-assisted workflow.

Get involved

Students: thesis / HiWi opportunities — see /student_thesis/ (or reach out to the team below).
Media: for expert interviews on inclusive AI, text simplification, and AI-assisted public communication, see /media/.

Lead: Michael Färber
Team: Gina Valentin; Daniel Spiering
Partners: Prof. Dr. Alexander Lasch; Claudia Neumann
Start: 09/2025 · Duration: 4 months

Text simplification is essential for making complex content accessible to diverse audiences who face comprehension challenges. Yet, the limited availability of simplified materials creates significant barriers to personal and professional growth and hinders social inclusion. Although researchers have explored various methods for automatic text simplification, none fully leverage large language models (LLMs) to offer tailored customization for different target groups and varying levels of simplicity. Moreover, despite its proven benefits for both consumers and organizations, the well-established practice of plain language remains underutilized. In this paper, we introduce https://simplifymytext.org, the first system designed to produce plain language content from multiple input formats, including typed text and file uploads, with flexible customization options for diverse audiences. We employ GPT-4 and Llama-3 and evaluate outputs across multiple metrics. Overall, our work contributes to research on automatic text simplification and highlights the importance of tailored communication in promoting inclusivity.

Updates

What we do

Why it matters

Get involved

References

2025