On Mon, May 4, 2015 at 11:13 PM, Julian Hyde <[email protected]> wrote:
> Your definition of a schema as a set of possible queries and the "can > be answered using" mathematical relation is a generalization of mine, > but I claim it is less useful. My lattice is defined structurally. It > only deals with filter-join-aggregate queries. Queries of this kind > are really important for interactive DW, and focusing on this kind of > queries makes the recommendation & matching algorithms tractable. > Actually, that wasn't my definition of a schema. That was my definition of the lattice in question. And I was expecting to restrict the discussion to the queries that you are talking about.
