Topology Analysis of Learning Cognitive Flow for Human-Computer Interaction

摘要

In recent years, improving the level of personalized education has attracted research interest of a great amount of people. However, it is difficult to make the high-dimensional cognitive learning space visible, computational and controllable. Whats more, manual data collection is difficult to meet the quantity and accuracy requirements, bringing obstacles in observing more accurate learning activities and collect more effective information. In this paper, a cognitive frame for observing the learning activities based on human-computer coupling is designed, for instance a vectorization technique for the situation of learning of human-computer is provided. Firstly, based on new cognitive philosophy such as cognitive distribution and extension, we propose a general topology of learning cognitive flow for human-computer interaction, composing of an evolving and high-dimensional system. The cognitive objects and their relationships are in an implicit learning, namely, cognitive space is in a human-computer coupling state. Those are distributed and extended to an information space, so as to form a "brain cognitive body-situation of coupling-manifold of information", which is a combination of cognition and information, named "BSM" structure. Furthermore, the mechanism for the BSM coupling morphism is analyzed, and the principle for the coupled observation of objects in a cognitive or learning manifold is proposed. At the same time, based on the concepts of category theory, such as commutative diagram and topology, a tree topology is selected as the topological structure of a low-dimensional learning space to process the observations of online learning. Finally, a special system for teaching is programmed to observe learning and training processes, thus summarizing knowledge points automatically to replace the manual way. The application of system will enable the teacher to better grasp students status of cognitive structure obviously, providing services to teaching. A new application framework for learning and a new idea for the scientific studies on learning are provided, supporting for "artificial education" to "human-computer learning" effectively.