增量数据驱动的自适应身份认证量化评估框架  

INCREMENTAL DATA-DRIVEN ADAPTIVE QUANTIFICATION EVALUATION FRAMEWORK FOR AUTHENTICATION METHODS

作　　者：李思斯[1] 韩伟力[1] Li Sisi;Han Weili(Software School,Fudan University,Shanghai 201203,China)

机构地区：[1]复旦大学软件学院

出　　处：《计算机应用与软件》2019年第11期294-300,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61572136);国家重点研发计划项目(2018YFC0830900)

摘　　要：准确评估身份认证风险是提高网络安全性的关键.当下访问控制机制缺乏对身份认证因子实用的量化评估方法.为此提出增量数据驱动的自适应身份认证量化评估框架,用于量化评估身份认证因子的安全性、易用性、可靠性、可访问性以及成本.采用基于专家经验的贝叶斯网络来综合评估因子组合的安全性;提出增量数据驱动的更新模型来根据实际运行时的认证记录动态调整量化评估结果.以某网站的认证方法分析为例,展示了该框架的合理性与可行性.Accurate evaluation of identity authentication risk is the key to improve network security.The current access control mechanism lacks a practical quantitative evaluation method for authentication factors.In this regard,we proposed an incremental data-driven adaptive authentication quantitative evaluation framework to quantify the security,usability,reliability,accessibility and cost of authentication factors.We used Bayesian network based on expert experience to evaluate the security of factor combination,and proposed an incremental data-driven update model to dynamically adjust the quantitative evaluation results according to the actual runtime authentication records.The rationality and feasibility of this framework are shown by the analysis of authentication methods of a website.

关 键 词：身份认证方法 量化评估 安全评估 数据驱动 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]