面向多主体的大学课程成绩相关性可视探索与分析  

Visual Exploration and Analysis of Multi-subject Correlation of Student Performance in College Courses

作　　者：纪连恩[1,2] 高芳 黄凯鸿 陈宗艳 Ji Lianen;Gao Fang;Huang Kaihong;Chen Zongyan(Department of Computer Science and Technology,China University of Petroleum,Beijing 102249;Beijing Key Laboratory of Petroleum Data Mining,Beijing 102249)

机构地区：[1]中国石油大学(北京)计算机科学与技术系,北京102249 [2]石油数据挖掘北京市重点实验室,北京102249

出　　处：《计算机辅助设计与图形学学报》2018年第1期44-56,共13页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(60873093)

摘　　要：深入分析学生成绩及其影响因素对于优化大学课程安排和提升教学质量具有重要意义.由于学生成绩数据涉及多个相互关联的分析主体,具有多元多属性和时序相关等特征,传统的分析工具和展示手段功能有限,难以有效探索影响课程成绩的多种关联因素,对异常现象也难以做出深入分析和解释.为此,文中根据成绩数据的特点设计了多视图协同交互的学生成绩可视分析系统——SPVAS.首先通过支持多维属性集成展示的矩阵热力图揭示学生成绩在年级和学期上的时序分布,其次为展示学生成绩中的多元统计特征以及相关联的课程和教师等主体特征,对平行坐标的交互展示能力进行扩充,最后为揭示课程成绩的影响因素以及课程间的相关性,设计弧长链接图与平行坐标和节点链接树相结合的创新布局,并应用多视图交叉筛选和动态关联等交互技术,实现从课程、学生和教师任意主体角度出发的交叉分析与连贯推理.为了验证原型系统的有效性和实用性,利用真实课程数据进行了案例研究,并邀请了相关的专业人员对本系统进行了试用与评价.In-depth analysis of student performance in college courses and its influence factors is significantto curriculum arrangement optimization and teaching quality improvement.However,it is challenging due to the complicated student score data which contains multiple relevant subjects and has characteristics like multivariate,multi-attribute and time-related.Traditional analysis tools and display methods are limitedwhen exploring association and anomalies in curriculum performance.In this paper,according to the char-acteristics of student score data,we design a student performance visual analysis system called SPVAS,which consists of multiple coordinated views.First1y,to explore the temporal distribution,variation patterns of student performance in different grades and semesters are discovered using heat-map matrix integrated with multi-attribute.Then,by extending interaction and demonstration ability of the parallel coordinates,multiple statistical characteristics of student performance and its correlative subjects such as courses and teachers can be presented.Finally,the correlation among courses and influence factors of student performance are revealed through novellayouts of combinations of arc diagram with both parallel coordinates and node-link diagram.With interaction techniques like cross filters and dynamic association in multiple views applying on the novellayouts,cross analysis and coherent inference from any aspects of course,student and teacher are achieved.To test the effectiveness and usefulness of SPVAS,a real dataset is adapted in a case study,and domain experts are involved during the process of test and evaluation.

关 键 词：大学课程成绩 多元多主体 相关性 弧长链接图 可视分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]