This leads to a perennial question: what is a collection? There is no bound on the number of answers to that. Some individual items may belong to multiple different collections. A herbarium catalogue, for example, could be split into species, or geographical coverage, or a host of other slices, all of which have valid meanings for particular searches. A Google search returns a “collection” of pages with common search text, but each of those pages may individually turn up in other searches (other “collections”). This is one of those questions where it best to think about the use of the data, how it will be found and accessed. At some point, aggregating every item into a single “collection” (‘everything I have ever measured’) does not add value to a search. Conversely, stripping every item out individually makes the complexity of managing it all much harder. If it is a problem in your application it may be useful to present multiple “collections” of the same underlying data, and let users find their own way into it. There is no good single answer.