readability-assessment
Filtering by topic readability-assessment(2)Clear all filters
- PaperarXiv — Language & NLP (cs.CL)20 Jun 2026
Can LLMs Control Readability? A Multi-Dimensional Evaluation Framework for CEFR-Controlled Arabic Generation
Nour Rabih, Chatrine Qwaider, Ted Briscoe
A multi-dimensional evaluation framework was proposed to assess LLMs' ability to control readability levels in Arabic text generation for CEFR levels. Structured prompting with lexical constraints significantly improved alignment, achieving high cosine similarity and near-perfect agreement with predicted readability levels.
Original abstract
While Large Language Models (LLMs) can generate fluent Arabic text, their ability to reliably control readability levels remains unclear. We propose a multi-dimensional evaluation framework for Common European Framework of Reference for Language (CEFR)-controlled Arabic text generation, assessing whether instruction-following LLMs can serve as reliable generators for adaptive language learning. Our framework integrates controlled prompting, automatic readability prediction using a validated Taha-19 model, lexical constraint validation, and syntactic complexity profiling. Results show that structured prompting substantially improves CEFR alignment. In particular, CEFR-guided prompting with lexical constraints achieves the highest conformity to reference linguistic profiles (0.91 cosine similarity) and near-perfect agreement with predicted readability levels (0.99), while unconstrained prompting exhibits weak control. These findings establish an empirical foundation for integrating readability-aware Arabic text generation into adaptive educational systems.
- PaperarXiv — Language & NLP (cs.CL)9 May 2026
Improving Lexical Difficulty Prediction with Context-Aligned Contrastive Learning and Ridge Ensembling
Wicaksono Leksono Muhamad, Joanito Agili Lopo, Tsamarah Rana Nugraha, Ahmad Cahyono Adi et al.
A Context-Aligned Contrastive Regression method integrating Ridge regression ensemble with contrastive objectives improves lexical difficulty prediction across different first-language backgrounds. The approach enhances cross-lingual representation alignment and captures ordinal difficulty structure, leading to more stable performance.
Original abstract
Lexical difficulty prediction is a fundamental problem in language learning and readability assessment, requiring models to estimate word difficulty across different first-language (L1) backgrounds. However, existing approaches rely on regression-only training with scalar supervision, which does not explicitly structure the representation space, limiting their ability to capture cross-lingual alignment and ordinal difficulty. To mitigate these issues, we propose Context-Aligned Contrastive Regression, which integrates Ridge regression ensemble with two complementary objectives, i.e., Cross-View Context and Ordinal Soft Contrastive Learning. Experiments on three L1 datasets show that (i) contrastive objectives improve cross-lingual representation alignment while preserving language-specific nuances, (ii) the learned representations capture the ordinal structure of lexical difficulty, and (iii) the ensemble effectively mitigates systematic biases of individual models, leading to more stable performance across difficulty levels.