Abstract
EDVIN is a subject term and genre data base for fiction and biographies, for use by librarians as well as by library users, based on many years of experience of what library users actually ask for, i.e., novels on a given subject, a novel similar to or as good as the one they have just read, or the title of a book on the basis of only vague recollections. The intention has been to find a way to answer those questions. The product is not a thesaurus constructed according to a previously-determined model, but a register of subject terms developed according to the a posteriori method and in close co-operation with the public.