Fuzzy Sets, Information Clouding and Database Security
Sujeet Shenoi · 1995
Database security is a novel and relatively unexplored area for applying fuzzy set theory. Fuzzy sets provide a powerful and appealing vehicle for clouding or “fuzzifying” sensitive information in databases. Their ability to incorporate natural semantics in clouded information helps balance user convenience and database security. In addition to controlling access to sensitive information, fuzzy sets can also cloud data dependencies. Such clouding helps enforce inference-control by eliminating precise reasoning involving sensitive information. Since snoopers can only reason approximately with clouded dependencies they can never infer exact confidential information. This paper describes a formal, readily implementable model for general-purpose relational database security. The model employs fuzzy sets in conjunction with a special context mechanism for clouding sensitive information maintained in database relations. The model supports a rich and secure query language. It is also possible to control inferences involving sensitive information by designing user views which only permit approximate or fuzzy reasoning.