探析深度学习表征的一种新方法:社会认知网络特征(SENS)

A New Method of Deep Learning Representation: Social Epistemic Network Signature(SENS)

近年来,深度学习在教育领域越来越受到重视,它被划分为三大维度:认知领域、人际领域和自我领域。已有的基于问卷、测试题和访谈等方式进行的学习效果分析,难以满足人们对于深度学习中的动态、过程和综合的分析需求。作为一种新的研究方法,社会认知网络特征(SENS)结合了社会网络分析(SNA)和认知网络分析(ENA)两种分析方法,能够很好地对深度学习的三个领域进行分析。通过对原有的SENS方法进行改进,提出了适合深度学习SENS方法的一般步骤,并对每个步骤进行详细的介绍;通过以两个典型案例展示如何利用SENS来促进深度学习,并将深度学习和在线协作数据分析相结合;提出了应用SENS的一些经验与建议,以期为今后研究深度学习提供新的思路和方法。

In recent years,deep learning has received more and more attention in the field of education,which is divided into three dimensions:cognitive,interpersonal and intrapersonal aspects.Current learning effect analysis are based on questionnaires,test questions and interviews,which are difficult to meet people’s demands for dynamic,procedural and comprehensive analysis in deep learning.As a new research method,social epistemic network signature(SENS)combines social network analysis(SNA)and cognitive network analysis(ENA)to analyze the three areas of deep learning.This study firstly modifies original SENS method,and puts forward the general steps of SENS method which was suitable for deep learning,with each step be explained in detail.Then,it introduces how SENS can be used to promote deep learning and combine deep learning with online collaborative data analysis through two typical cases.Finally,some experience and suggestions on the application of SENS are given in order to provide new ideas and methods for future researchers to carry out deep learning research.