An "information universe" is defined by its indexing

An information source can be logically defined by the actual data content and the method(s) that can be applied to access that data, i.e., semantics counts.

Two sets of data are said to be in different universes if the methods used to access one are meaningless in the other.

For instance, a chemical catalog such as ACD (indexed by catalog number, catalog/preferred names, chemical structure) and PubMed/Entrez (indexed by author, journal, and search term) are in different information universes because a query which is meaningful for one is not meaningful in the other.