The notion of information content within a database-oriented system has proven to be an elusive one. This paper expands upon an earlier definition of information content, and explores in depth the use of the information carried by data in a database. From the extended definition, a definition of “information content inclusion relation (IIR)” is derived. The paper then formulates “inference rules for reasoning logically about such relations.”
A key concept presented is that information content is different than the linguistic meaning of a data construct. The paper reasons about how this difference “might be exploited with a database.” Information content may be used to derive otherwise hidden information (that standard query tools would not be able to find). A prototype analysis tool is described that demonstrates how to take advantage of IIR relationships within databases.
Database management system (DBMS) academics and vendor architects might find this paper useful. However, readers must be willing to invest a good deal of time reviewing the formal mathematical reasoning.