Mining sequential patterns allows the analysis of large data-
bases of sales data that are not susceptible for investigation ''by hand''.
Unfortunately, data mining algorithms produce a large number of re-
sults, that often exceed the size of the original database. Analysis of data
mining results is usually performed by the end user of the data mining
application manually. During this process a large number of subsequence
queries are executed. Such queries are not well supported by traditional
database management systems. In this paper we present a novel inde-
xing technique for sequences of sets such as sequential patterns or sales
history. Experimental evaluation of the index proves the feasibility and
benefit of the index in exact and similar matching of subsequences.