Providing appropriate learning partners to support collaborative learning is crucial to mitigate distance learners' aloneness in learning process and improve the learning effect. This paper proposes a novel six-dimension learning behavior model based on the authentic data collected from the network learning platform. By employing the similarity algorithm, the model can distinguish the similar behavior model or complementary behavior model. As a result, learners can either receive the autonomously recommended learning partners or choose their own preferred partners. Experiment results verified the validity of the proposed learning behavior model and the recommendation algotithms. Compared with traditional grouping models, this model-supported learning partner recommendation is more personalized and intelligent, and is more suitable in large-scale network learning environment.%如何提供合适的学习伙伴进行协作学习,是解决远程学习者学习孤独、提升学习效果的关键问题.文章通过网络教学平台采集数据,提出并建立了六维的学习者学习行为模型,并利用相似度区分相似型学习行为模型和互补型学习行为模型,进而为学习者推荐相似型学习伙伴和互补型学习伙伴,而学习者也可以自主选择学习伙伴.实验结果验证了学习行为模型和推荐方法的有效性.相比已有的分组方式,基于学习行为模型的学习伙伴推荐方法具有可视化、智能性、个性化等特点,更适合大规模网络学习环境下的学习伙伴推荐.
展开▼
