Abstract interpretation of database query languages

In this paper, we extend the Abstract Interpretation framework to the field of query languages for relational databases as a way to support sound approximation techniques. This way, the semantics of query languages can be tuned according to suitable abstractions of the concrete domain of data. The abstraction of relational database system has many interesting applications, in particular, for security purposes, such as fine grained access control, watermarking, etc.

► A new concrete/abstract denotational semantics of SQL embedded applications.
► A basis for Abstract Interpretation-based analyses for programs embedded with SQL.
► Supports fine-grained access control, digital watermarking, integrity constraints checking.