German4All - A Dataset and Model for Readability-Controlled Paraphrasing in German

2025-08-25

AI Summary

This study addresses controllable readability paraphrasing for German text, targeting readers across multiple proficiency levels. Methodologically, we introduce the first large-scale, five-level aligned German controllable readability rewriting dataset (25,000 samples), synthesized via GPT-4 and rigorously validated through human annotation and LLM-based quality assessment. Leveraging this dataset, we train a deep learning model enabling fine-grained control over discrete readability levels. Our key contributions are: (1) releasing the first open-source, multi-level German readability rewriting benchmark; (2) open-sourcing a high-performance controllable paraphrasing model; and (3) achieving state-of-the-art performance on German text simplification. The proposed approach significantly enhances textual accessibility for diverse audiences—including children, second-language learners, and individuals with cognitive impairments—by enabling precise, level-aware simplification.

Technology Category

• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Humans and AI: AI for Accessibility
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

The ability to paraphrase texts across different complexity levels is essential for creating accessible texts that can be tailored toward diverse reader groups. Thus, we introduce German4All, the first large-scale German dataset of aligned readability-controlled, paragraph-level paraphrases. It spans five readability levels and comprises over 25,000 samples. The dataset is automatically synthesized using GPT-4 and rigorously evaluated through both human and LLM-based judgments. Using German4All, we train an open-source, readability-controlled paraphrasing model that achieves state-of-the-art performance in German text simplification, enabling more nuanced and reader-specific adaptations. We opensource both the dataset and the model to encourage further research on multi-level paraphrasing

Problem

• Creating readability-controlled paraphrases for German texts
• Addressing lack of large-scale German paraphrasing datasets
• Enabling tailored text adaptations for diverse reader groups

Innovation

• GPT-4 synthesized German readability dataset
• Open-source readability-controlled paraphrasing model
• Multi-level text simplification for diverse readers