Data abstraction tools
Joachim W. Schmidt · ACM SIGART Bulletin · 1980
Current research in data modeling is motivated by the following dilemma: - At the application level - being confronted with “slices of reality” - details are perceived that, in general, cannot be represented. - At the representation level - being confronted with “levels of machines” - details are represented that, in general, cannot be perceived. Abstraction methods cope with that problem by suppressing unnecessary details and by formalizing and structuring the relevant information.