在线学习风格识别技术与方法研究  

Research on Online Learning Style Recognition Techniques and Methods

作　　者：黄诚[1] 张冰雪 赵逢禹[1] HUANG Cheng;ZHANG Bing-xue;ZHAO Feng-yu(School of Optical-Electrical and Computer Engineering,University of Shanghai for Science and technology,Shanghai 200093 China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《软件》2018年第4期97-102,共6页Software

摘　　要：每个学习者在学习过程中都有自己的学习风格。现有的利用学习行为自动识别学习风格的方法所采用的行为数据单一,未能充分体现Felder–Silverman理论中与学习风格相对应的学习行为的定义,且大部分学习风格识别模型都只采用单一算法对行为特征数据进行处理,识别效率不高。针对这两个问题提出了一种利用每个风格维度最少4种在线行为特征数据并结合C4.5决策树算法和隐马尔可夫算法识别学习风格的方法,并比较每种算法对4种不同学习风格维度识别的优劣程度。通过对Moodle进行二次开发获取数据进行实验,实验验证了该方法的有效性。Each learner has his own learning style in the learning process. The existing single behavioral data used in the method of learning style recognition automatically by learning behavior can not fully reflect the definition of learning behavior corresponding to learning style in Felder-Silverman theory, and most of the learning style recognition models use only a single algorithm to processes the behavioral characteristic data and has low recognition efficiency. Aiming at these two problems, a method of using at least 4 kinds of online behavioral data for each style dimension and combining C4.5 decision tree algorithm with Hidden Markov algorithm to identify learning styles and compares the advantages and disadvantages of each algorithm for recognizing the dimensions of four different learning styles. Through the secondary development of Moodle data acquisition experiment, the experimental verification of the effectiveness of the method.

关 键 词：学习风格模型 行为特征 C4.5决策树 隐马尔可夫模型 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]