基于认知诊断理论的网络安全自适应测试技术
|
摘要
为进一步研究人员的网络安全素养,准确诊断人员网络安全知识和技能的水平,结合心理测量学和计算机测试技术,开发了基于认知诊断的多级属性评分的自适应测试技术。首先,为更好适应多元化复杂的网络安全知识结构且便于测试模型,在网络安全领域设计了复杂的层级网络安全知识库模型;然后,在多级评分认知诊断模型的基础上引入了属性层级的概念进行综合改进,并提出了准确、高效的参数估计方法和同模型相适应的选题策略。实验结果表明,多级属性评分的网络安全自适应测试技术较传统的多级评分模型提高了10.5%的效率,为计算机自适应测试领域的研究提供了参考。
To further effectively study the cybersecurity literacy of personnel and accurately diagnose the specific level of personnel knowledge and skills,the paper developed adaptive cybersecurity testing technology based on multi-level attributes scoring of cognitive diagnosis by combining psychometrics and computer testing technology.Firstly,a hierarchical cybersecurity knowledge model is designed for better adapting to the complex knowledge structure and verifying the research.Then,the hierarchy of attribute is input based on polytomous scoring cognitive diagnosis model to implement comprehensive improvements.Accurate and efficient parameter estimation method and suitable selection strategy are proposed to improve performance.The experimental results show that the adaptive cybersecurity testing technology of multi-level attributes scoring improves the efficiency by 10.5% compared with the traditional multi-level scoring model,which provides a reference to the research of computerized adaptive testing.
To further effectively study the cybersecurity literacy of personnel and accurately diagnose the specific level of personnel knowledge and skills,the paper developed adaptive cybersecurity testing technology based on multi-level attributes scoring of cognitive diagnosis by combining psychometrics and computer testing technology.Firstly,a hierarchical cybersecurity knowledge model is designed for better adapting to the complex knowledge structure and verifying the research.Then,the hierarchy of attribute is input based on polytomous scoring cognitive diagnosis model to implement comprehensive improvements.Accurate and efficient parameter estimation method and suitable selection strategy are proposed to improve performance.The experimental results show that the adaptive cybersecurity testing technology of multi-level attributes scoring improves the efficiency by 10.5% compared with the traditional multi-level scoring model,which provides a reference to the research of computerized adaptive testing.
引文
[1] CONTEH N Y,SCHMICK P J.Cybersecurity:risks,vulnerabi- lities and countermeasures to prevent social engineering attacks[J].International Journal of Advanced Research in Computer Science,2016,6(23):31-38.
[2] GRATIAN M,BANDI S,CUKIER M,et al.Correlating Human Traits and Cybersecurity Behavior Intentions[J].Computers & Security,2018,73(3):345-358.
[3] BASSETT G.System and method for cyber security analysis and human behavior prediction:US 20160205122.A1[P].2016-3-22.
[4] YOUNG H,VLIET T V,VEN J V D,et al.Understanding Human Factors in Cyber Security as a Dynamic System[C]//AHFE 2017:8th International Conference on Applied Human Factors and Ergonomics.Los Angeles,Springer,2017:244-254.
[5] ZHANG H L,YU H N,FANG B X,et al.Research on China’s cyberspace security practice qualification system [J].Chinese Engineering Science,2016,18(6):44-48.(in Chinese)张宏莉,于海宁,方滨兴,等.我国网络空间安全执业资格认证体系研究[J].中国工程科学,2016,18(6):44-48.
[6] SMITS N,PAAP M C S,B?HNKE J R.Some recommendations for developing multidimensional computerized adaptive tests for patient-reported outcomes[J].Quality of Life Research,2018,27(4):1055-1063.
[7] GU Y,XU G.The Sufficient and Necessary Condition for the Identifiability and Estimability of the DINA Model[J].Psychometrika,2018(2):1-16.
[8] TORRE J D L,MINCHEN N.Cognitively Diagnostic Assessments and the Cognitive Diagnosis Model Framework[J].Psicología Educativa,2014,20(2):89-97
[9] MCS P,KROEZE K A,TERWEE C B,et al.Item usage in a multidimensional computerized adaptive test (MCAT) measu-ring health-related quality of life[J].Quality Life of Research,2017,26(11):2909-2918.
[10] RAJ R K,PARRISH A.Toward Standards in Undergraduate Cybersecurity Education in 2018[J].Computer,2018,51(2):72-75.
[11] QI B,WANG Y,ZOU H X,et al.The Analysis of Measurement Method in the Knowledge System of Network Security Based on Information Entropy[C]//ICCT 2017:17th IEEE International Conference on Communication Technology.Sichuan,China:IEEE Press,2017:1328-1333.
[12] TU D B,CAI Y,DAI H Q,et al.A Polytomous Cognitive Diagnosis Model:P-DINA Model[J].Acta PsychologicaSinica,2010,42(10):1011-1020.(in Chinese)涂冬波,蔡艳,戴海琦,等.一种多级评分的认知诊断模型:P-DINA模型的开发[J].心理学报,2010,42(10):1011-1020.
[13] CAI Y,MIAO Y,TU D B.The polytomously scored cognitive diagnosis computerized adaptive testing[J].Acta PsychologicaSinica,2016,48(10):1338-1346.(in Chinese)蔡艳,苗莹,涂冬波.多级评分的认知诊断计算机化适应测验[J].心理学报,2016,48(10):1338-1346.
[14] XU G.Identifiability of restricted latent class models with binary responses[J].The Annals of Statistics,2017,45(2):675-707.
[15] HSU C L,WANG W C,CHEN S Y.Variable-Length Computerized Adaptive Testing Based on Cognitive Diagnosis Models.[J].Applied Psychological Measurement,2013,37(7):563-582.
[16] KAPLAN M,TORRE J D L,BARRADA J R.New Item Selec- tion Methods for Cognitive Diagnosis Computerized Adaptive Testing[J].Applied Psychological Measurement,2015,39(3):167-188.
[2] GRATIAN M,BANDI S,CUKIER M,et al.Correlating Human Traits and Cybersecurity Behavior Intentions[J].Computers & Security,2018,73(3):345-358.
[3] BASSETT G.System and method for cyber security analysis and human behavior prediction:US 20160205122.A1[P].2016-3-22.
[4] YOUNG H,VLIET T V,VEN J V D,et al.Understanding Human Factors in Cyber Security as a Dynamic System[C]//AHFE 2017:8th International Conference on Applied Human Factors and Ergonomics.Los Angeles,Springer,2017:244-254.
[5] ZHANG H L,YU H N,FANG B X,et al.Research on China’s cyberspace security practice qualification system [J].Chinese Engineering Science,2016,18(6):44-48.(in Chinese)张宏莉,于海宁,方滨兴,等.我国网络空间安全执业资格认证体系研究[J].中国工程科学,2016,18(6):44-48.
[6] SMITS N,PAAP M C S,B?HNKE J R.Some recommendations for developing multidimensional computerized adaptive tests for patient-reported outcomes[J].Quality of Life Research,2018,27(4):1055-1063.
[7] GU Y,XU G.The Sufficient and Necessary Condition for the Identifiability and Estimability of the DINA Model[J].Psychometrika,2018(2):1-16.
[8] TORRE J D L,MINCHEN N.Cognitively Diagnostic Assessments and the Cognitive Diagnosis Model Framework[J].Psicología Educativa,2014,20(2):89-97
[9] MCS P,KROEZE K A,TERWEE C B,et al.Item usage in a multidimensional computerized adaptive test (MCAT) measu-ring health-related quality of life[J].Quality Life of Research,2017,26(11):2909-2918.
[10] RAJ R K,PARRISH A.Toward Standards in Undergraduate Cybersecurity Education in 2018[J].Computer,2018,51(2):72-75.
[11] QI B,WANG Y,ZOU H X,et al.The Analysis of Measurement Method in the Knowledge System of Network Security Based on Information Entropy[C]//ICCT 2017:17th IEEE International Conference on Communication Technology.Sichuan,China:IEEE Press,2017:1328-1333.
[12] TU D B,CAI Y,DAI H Q,et al.A Polytomous Cognitive Diagnosis Model:P-DINA Model[J].Acta PsychologicaSinica,2010,42(10):1011-1020.(in Chinese)涂冬波,蔡艳,戴海琦,等.一种多级评分的认知诊断模型:P-DINA模型的开发[J].心理学报,2010,42(10):1011-1020.
[13] CAI Y,MIAO Y,TU D B.The polytomously scored cognitive diagnosis computerized adaptive testing[J].Acta PsychologicaSinica,2016,48(10):1338-1346.(in Chinese)蔡艳,苗莹,涂冬波.多级评分的认知诊断计算机化适应测验[J].心理学报,2016,48(10):1338-1346.
[14] XU G.Identifiability of restricted latent class models with binary responses[J].The Annals of Statistics,2017,45(2):675-707.
[15] HSU C L,WANG W C,CHEN S Y.Variable-Length Computerized Adaptive Testing Based on Cognitive Diagnosis Models.[J].Applied Psychological Measurement,2013,37(7):563-582.
[16] KAPLAN M,TORRE J D L,BARRADA J R.New Item Selec- tion Methods for Cognitive Diagnosis Computerized Adaptive Testing[J].Applied Psychological Measurement,2015,39(3):167-188.