[1]唐宇坤,邓 松*,许梦雅,等.基于几何特征的学生评教数据离群点检测算法[J].江西师范大学学报(自然科学版),2021,(03):292-298.[doi:10.16357/j.cnki.issn1000-5862.2021.03.11]
　TANG Yukun,DENG Song*,XU Mengya,et al.The Outlier Detection Algorithm for Student Evaluation Data Based on Geometric Features[J].Journal of Jiangxi Normal University:Natural Science Edition,2021,(03):292-298.[doi:10.16357/j.cnki.issn1000-5862.2021.03.11]
点击复制

基于几何特征的学生评教数据离群点检测算法()
分享到：

《江西师范大学学报》（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2021年03期
页码:
292-298
栏目:
信息科学与技术
出版日期:
2021-06-10

文章信息/Info

Title:
The Outlier Detection Algorithm for Student Evaluation Data Based on Geometric Features
文章编号:
1000-5862(2021)03-0292-07
作者:
唐宇坤; 邓 松*; 许梦雅; 郭 馨
江西财经大学软件与物联网工程学院,江西 南昌 330013
Author(s):
TANG Yukun; DENG Song*; XU Mengya; GUO Xin
College of Software and Internet of Things Engineering,Jiangxi University of Finance and Economics,Nanchang Jiangxi 330013,China
关键词:
学生评教;  几何特征;  离群点;  支持向量机
Keywords:
student evaluation;  geometric features;  outliers;  support vector machine
分类号:
TP 311
DOI:
10.16357/j.cnki.issn1000-5862.2021.03.11
文献标志码:
A
摘要:
针对学生评教数据中的离群点问题,根据消极评教数据产生的方式及特点,提出了一种基于几何特征的学生评教数据离群点检测算法.该算法通过分析样本的几何特征,计算样本的离群程度,完成离群点检测,共分为3步进行:(i)依据教学质量评价数据,在几何特征空间中建立样本的点映射;(ii)从形状相似度、距离相似度2个方面构建判别空间,对几何特征空间中的样本点进行分析运算,得到样本点在判别空间中的点映射;(iii)以基于半监督近邻的方法对判别空间中的样本进行检测.实验结果表明:该算法检测精度较高,在高校教师教学效果中有较好的应用价值.
Abstract:
To solve the outlier problem in the student evaluation data,a outlier detection algorithm for the students' evaluation data based on geometric feature is presented according to the way and characteristics of the data that does not fit in with the evaluation data.By analyzing the geometric characteristics of the samples,the algorithm calculates the outlier degree of the samples and completes the outlier detection,which is divided into three steps.Firstly,based on the teaching quality evaluation data,the point mapping of samples is established in the geometric feature space.Secondly,the discriminant space is constructed from the shape similarity and distance similarity.The point mapping of sample points in the discriminant space is obtained by analyzing and calculating the sample points in the geometric feature space.Finally,the samples in the discriminant space are tested based on semi-supervised neighbors.The experimental results show that the algorithm has a high detection accuracy and has a good application value in the teaching effect of university teachers.

参考文献/References:

[1] 韩映雄,周林芝.学生评教的信度、效度、影响因素及应用风险[J].复旦教育论坛,2018,16(6):74-81.
[2] 徐雪珂,邓松,张荣,等.基于个体特征与教学评价的教师学习对象推荐[J].江西师范大学学报:自然科学版,2019,43(4):409-415.
[3] Aggarwal C C.Outlier analysis[M].2nd ed.Cham:Springer,2017:286-286.
[4] Zhi Hongzhi,Bah M J,Hammad M.Progress in outlier detection techniques:a survey[J].IEEE Access,2019,7:107964-108000.
[5] Xu Xiaodan,Liu Huawen,Yao Minghai.Recent progress of anomaly detection[EB/OL].[2020-02-19].https://downloads.hindawi.com/journals/complexity/2019/2686378.pdf.
[6] 吴镜锋,金炜东,唐鹏.数据异常的监测技术综述[J].计算机科学,2017,44(S2):24-28.
[7] 董秋仙,朱赞生.一种新的选取初始聚类中心的k-means算法[J].统计与决策,2020,36(16):32-35.
[8] 王子龙,李进,宋亚飞.基于距离和权重改进的k-means算法[J].计算机工程与应用,2020,56(23):87-94.
[9] 张巍,王洋,刘东宁,等.基于随机聚类方法建模的序列分析[J].江西师范大学学报:自然科学版,2017,41(5):470-475.
[10] 何慧,胡小红,覃华,等.用核k-means聚类减样法优化半定规划支持向量机[J].江西师范大学学报:自然科学版,2013,37(6):574-578.
[11] 仇开,姜瑛.加权LOF结合上下文判断的云环境中服务运行数据异常检测方法[J].计算机工程与科学,2020,42(6):951-958.
[12] 贺寰烨,林果园,顾浩,等.云虚拟机异常检测场景下改进的LOF算法[J].计算机工程与应用,2020,56(23):80-86.
[13] 梅林,张凤荔,王瑞锦,等.基于网格划分加权的分布式离群点检测算法[J].电子科技大学学报,2020,49(6):860-866.
[14] 赵颖,哈巍.分数膨胀的影响因素与学生选课策略[J].清华大学教育研究,2020,41(6):63-74.
[15] 蒋贵友,郭丽君.大学评教“共谋”行为及其治理路径[J].大学教育科学,2020(2):105-110.

相似文献/References:

[1]唐宇坤,邓 松*,唐熙淳,等.基于矛盾关系的评教文本反语检测算法[J].江西师范大学学报(自然科学版),2022,(01):59.[doi:10.16357/j.cnki.issn1000-5862.2022.01.08]
　TANG Yukun,DENG Song*,TANG Xichun,et al.The Irony Detection Algorithm of Student's Teaching Evaluation Data Based on Contradiction[J].Journal of Jiangxi Normal University:Natural Science Edition,2022,(03):59.[doi:10.16357/j.cnki.issn1000-5862.2022.01.08]

备注/Memo

备注/Memo:
收稿日期:2020-12-17
基金项目:国家自然科学基金(61462037)资助项目.
通信作者:邓 松(1982—),男,江西南昌人,副教授,博士,主要从事实体关联、数据法学和教育信息化研究.E-mail:47817086@qq.com
更新日期/Last Update: 2021-06-10