lexical-difficulty

This paper introduces Context-Aligned Contrastive Regression to improve lexical difficulty prediction by addressing cross-lingual alignment and ordinal structure challenges in language learning datasets.