Note on:

     "Discovery of frequent episodes in event sequences",
     Heikki Mannila, Hannu Toivonen, A. Inkeri Verkamo,
     Department of Computer Science,
     University of Helsinki, Finland,
     Proceedings of The First International Conference on 
     Knowledge Discovery and Data Mining (KDD 1995), pp 210-215

This paper deals with finding episodes, collections of events, that
occur inside a window of time specified by size. Events in this
context describes an item occurring in a single point of a sequence. 

If I were to consider the beginning and ending of each of my events as
a single event it would make an interesting comparison. There are a
few obstacles to doing this. I do not have the constraint of windows
though the concept could be easily added to Apriori Sets And
Sequences. I have normal items in my data set which do not occur at
any particular point in time. Items of this sort are not represented
in this paper.

Another subtle problem in comparing Apriori Sets And Sequences and
this work is that the input here is a single sequence. The input to
ASAS consists of many instances containing potentially many sequences
in addition to regular, single valued attributes.

The authors say something that strikes me a bit odd. They briefly
describe the Apriori prune and cite Agrawal and Srikant 1994, as well
as another paper they authored as describing similar ideas. It seems to
me that the ideals are the same.