<![CDATA[AI-driven vocabulary recommender systems can personalize vocabulary practice by using learner data such as prior performance, error patterns, goals, and review schedules. However, many systems remain opaque because they recommend words without explaining why the items are selected or how learners can adjust the recommendation process. This conceptual article examines transparency as a key design issue in technology-enhanced language learning and synthesizes literature on explainable AI, trust in automation, technology acceptance, and self-regulated learning. The results of the synthesis show that transparent vocabulary personalization is most likely to support learning when it strengthens three mechanisms: perceived control, calibrated trust, and sustained adoption. The analysis also indicates that the most appropriate design is not excessive technical disclosure, but concise explanations, on-demand details, and learner controls that allow users to correct or adjust recommendations. These findings suggest that trustworthy vocabulary recommenders should combine pedagogically meaningful explanations with learner agency so that adaptive systems support, rather than replace, self-regulated learning.]]>
Copyrights © 2026
