Hierarchical Storage in Information Retrieval
A probabilistic analysis is employed to determine
the effect of hierarchical storage organizations 
on information retrieval operations.  The data storage
hardware is assumed to consist on n-levels of 
linearly connected memory hardware with increasing data
access times and increasing data storage capabilities. 
 A system might, for example, consist of fast semiconductor
memory, computer core memory, extended core 
storage, disk memory, and data cells.  Equations are derived
to predict the effect of such a system on 
data access times using sequential files, random access files,
and structured files employing multiple-hierarchical 
linked lists.
CACM May, 1973
Salasin, J.
