#1242: CoReTra: Supporting Learners in Identifying Personally Meaningful Contrasts among Readers’ Interpretations

Writing that involves personally reconstructing what one has read is essential for knowledge transformation, although university students often struggle to make meaningful connections between readings and their own ideas. Existing instructional approaches offer procedural guidance but often do not support deep, personally meaningful integration. To address this issue, we developed CoReTra (Collaborative Reflection through conTrast of interpretations), a web-based system designed to support learners in identifying and integrating personally meaningful contrasts. The system includes two views: the Decomposition View, which allows learners to extract content elements from readings and record notes as cards, and the Integration View, which enables learners to compare cards from different perspectives to refine their interpretations. The system uses large language model (LLM)–based semantic mining to adaptively suggest peer-created cards that represent diverse viewpoints. These features support learners in progressively revisiting and reconstructing their interpretations, promoting deeper reflection in CSCL contexts, while keeping contrast identification learner‑driven.