基于多标记的多样性学习特征分类框架

A Diverse Learning Feature Classification Framework Based on Multi-label

网络学习的最终目的是让学习者掌握隐性知识及其建构方法,如何对学习者隐性知识的建构能力进行评测成为研究网络学习方法的关键问题。提出一种对模式知识进行学习分析的多样性学习特征分类框架。在该框架中首先将隐性知识映射到模式知识,然后依据学习者模式知识的学习状态采用聚类分析方法进行分析,接着在学习评测中采用多标记学习算法对学习特征进行分类,最后选择5种常用的评价指标衡量测评结果。实验结果表明,该多样性学习特征分类框架在5种常用评价指标上都取得了较好的分类效果,采用该框架有利于分析学习者的学习能力。

The ultimate goal of online learning is to enable learners to master tacit knowledge and its construction methods.How to evaluate learners ability to construct tacit knowledge has become a key issue in researching online learning methods.This paper proposes a diverse learning feature classification framework for learning and analyzing pattern knowledge.In this framework,tacit knowledge is first mapped to pattern knowledge,and then clustering analysis is carried out based on the learning status of learners pattern knowledge.Then,the learning features are classified by multi-label learning algorithm in learning evaluation,and finally five commonly used evaluation indicators are selected to measure the evaluation results.The experimental results show that the diverse learning feature classification framework achieves good classification performance on five commonly used evaluation indicators.The adoption of the framework is beneficial for analyzing learners learning ability.