Temporal tracking and early warning of multi semantic features of learning behavior

Abstract

From the perspective of computer-supported learning, the temporal tracking and early warning plays an important role, and it is also an effective means to improve learning interest and learning behavior. In this study, based on massive and complete data, the temporal tracking and early warning of multi semantic features are analyzed and designed. Firstly, the multidimensional factors are explored, and the semantic features and relationships are defined. Secondly, based on the complexity and particularity of learning behavior, a multi semantic convolution network prediction model (iCaNN) is designed. Through sufficient model training and testing, the optimal experimental configuration is demonstrated; Thirdly, we visualize the multi semantic features, mine the existed problems, and explore the key learning rules and temporal risks. Finally, the effectiveness of temporal tracking and early warning is tested in practice. The whole research is of great significance to improve learning behavior, and bring positive learning effectiveness. • The multidimensional influencing factors of learning process are explored. • The semantic features and relationships are defined. • A multi semantic convolution network prediction model is designed. • The key learning rules and temporal risks is mined and explored. • The effectiveness of temporal tracking and early warning is tested in practice.