Vocabulary acquisition is the foundation of language learning, and Computer-Assisted Vocabulary Learning (CAVL) environments offer new possibilities for enhancing learning efficiency. However, conventional CAVL environments often employ static learning content and limited interaction mechanisms, making it difficult to accommodate individual learner differences and cognitive needs, which leads to insufficient motivation and suboptimal memory retention. This study proposed an Intelligent Interactive Companion-assisted Vocabulary Learning (IIC-VL) approach, which was based on multi-agent technology and dynamically adjusted learning content according to learners' proficiency levels, learning styles, and emotional states. An experimental design was adopted, with 100 students recruited from a high school and randomly assigned to an experimental group (n = 49) and a control group (n = 51). The experimental group used the IIC-VL approach for vocabulary learning, while the control group used the conventional CAVL approach. The intervention lasted for 12 wk, with two learning sessions per week, each lasting 45 min. The results showed that, compared to the control group, the experimental group demonstrated significant improvements in vocabulary tests (both receptive and productive), motivation for learning, and flow experience. Subsequent semi-structured interviews revealed students' positive attitudes and perceptions toward the IIC-VL approach. These findings not only provided empirical support for the application of multi-agent technology in CAVL environments but also offered a new practical direction for personalized learning paths in foreign language vocabulary instruction.
Zhang et al. (2025) studied this question.