姜 强， 药文静， 赵 蔚， 李 松
【关 键 词 】：
深度学习； 动态知识图谱； 协同知识建构； ARCS模型； 文本挖掘； 滞后序列分析
Deep learning, which responds to the epoch demands and the reform of "core literacy", answers the question of "what kind of people to cultivate" and returns to the essence of learning. Knowledge map can help students to think deeply, improve their problem solving ability, critical thinking and creative ability as well, and realize deep learning. However, the previous knowledge construction had the limitation of static and isolated organization. Based on ARCS motivation model and knowledge construction theory, a model to construct the dynamic knowledge map for deep learning is built from collaborative knowledge construction, motivational strategies and learning environment, which is characterized by dynamic generation, timely feedback and interactive sharing, so as to highlight students' subjectivity and initiative and enhance their learning experience. Taking college students as the research object, this paper analyzes the effect of the model by means of text mining and lag sequence analysis. The results show that the experimental group is superior to the control group in terms of academic performance, attention, etc. Particularly, this model has a positive effect on learners with middle-level or low-level motivation, and their perceived attention, confidence and satisfaction in completing tasks have been significantly improved. The construction and development of dynamic knowledge map can be explored from the aspects of remolding pre-task planning, social cognitive openness, meaning negotiation and generative teaching, so as to promote deep cognitive ability and higher-order thinking.