Text difficulty prediction in educational contexts requires models that balance predictive performance, interpretability, calibration, and pedagogical alignment. While transformer-based approaches increasingly dominate text difficulty classification, educational applications demand transparent and linguistically grounded modeling. This paper presents work aimed at developing a workbench for CEFR-based text difficulty prediction. The proposed platform comprises three main components: (i) a tool for CEFR-aligned dataset preparation incorporating a pipeline for documenting, processing, and enriching textual data, (ii) CEFR-aligned datasets, and (iii) three alternative modeling approaches, namely a rule-based baseline, a feature-based Machine Learning (ML) classifier, and a fine-tuned BERT model. Our approach integrates linguistically informed feature engineering with data-driven modeling techniques, thereby balancing transparency and predictive performance. The proposed workbench has been designed as a language-agnostic infrastructure that can be extended to any language. In its current implementation, it has been applied to the creation of a German CEFR dataset, while its Greek counterpart is currently under development.

Foreign language learning is a cognitively and affectively demanding process, in which fluctuations in attention and motivation can negatively impact learner engagement. Emotions play a central role in this process, yet they are rarely modelled in a systematic, data-driven manner in authentic learning environments. At the same time, positive emotions. This paper presents a prototype affective computing architecture that incorporates various modalities (audio, video, biosignals) to facilitate real-time or near-real-time emotion recognition in an educational scenario; the architecture is integrated within an emotion-aware and adaptive Language Learning application that harnesses Large Language Models in view of providing appropriate educational scenarios to learners. The system comprises modules for acquiring data for each modality and a processing pipeline for synchronizing and analyzing heterogeneous affective signals. We demonstrate both the feasibility and applicability of the approach through a proof-of-concept implementation and discuss its relevance for studying learner affect and supporting affect-aware educational scenarios. The results highlight both the applicability of multimodal affective data in educational settings and the need for further research on their pedagogical interpretation and use.