【摘要】 数据开放为海量数据中的价值得以最大化利用提供了可能,然而数据的安全性问题却成开放共享的最大阻碍。针对开放数据集查询分析结果包含的隐私信息,基于其与数据访问行为的直接联系,设计一种安全规则存储结构。提出面向自然语言的安全需求描述接口,对待保护的隐私信息进行灵活、方便的描述,进一步提出自然语言隐私需求描述到安全规则的自动转换方法。在此基础上建立数据安全自动审核模型,该模型根据数据拥有者的安全需求对访问者的数据行为进行审核,实现在数据隐私不泄露的前提下保证数据开放的最大化。通过真实数据集上的实验表明:安全规则能够准确地捕获数据提供者的隐私保护需求,并能够有效地保证数据的安全性。更多还原

【Abstract】 Data openness provides the great opportunities to maximize the value in massive data,but the data security issue becomes the big obstacle to open sharing. In view of the privacy information contained in the results of open data set query analysis,a storage structure of the security rules was designed based on its direct connection with data access behavior. Natural language-oriented security requirements description interface was proposed to describe the protected privacy information flexibly and conveniently. Furthermore,an automatic conversion method from privacy requirement description of natural language to security rules was proposed. On this basis,an automatic data query verification model was established,which checked the data behavior of users according to the security needs of data owners. It realized the maximization of data openness without revealing data privacy. Experiments on real data sets show that security rules can accurately capture the privacy protection needs of data providers and effectively guarantee the security of data.更多还原