Probably the first classification exercise in your life: sorting your toys into boxes.
The names on each box have to change as your collection of toys grows.

"Models that accurately handle taxonomic data are inherently complex ..." (Pullan et al. 2000).

The majority of taxonomic databases take a rather simple view of taxonomy and demand arbitration of taxonomic opinion in their construction. Here we describe a system designed to allow the construction of a nomenclatural database which provides an audit trail for the history of any given name. Conceptually, it mimics the taxonomist's notebook while engaged in nomenclatural research in the library.

The key advantage of this system is that it is maintenance-free, in the sense that data, once entered, are never changed except to correct entry errors. New data are added cumulatively and the required taxonomic view-point is constructed as part of the query process, rather than being inherent in the database structure.