Download Document

Mining Temporal Patterns for Interval-Based and Point-Based
Events
Abstract:
Preceding research on mining sequential patterns mainly focused on
determining patterns from point-based event data and interval–based event data,
where a pair of time values is associated with each event. Since many areas of
research
includes
data
on
a
snapshot
of
time
points
as well as time intervals, it is necessary to define a new temporal pattern. In this
work, based on the existing thirteen temporal relationships, a new altenative of
temporal pattern is defined for interval based as well as point–based event data.
Then, a hybrid pattern mining technique is proposed.Experimental results show that
the extensiveness and precision of the proposed hybrid technique are more powerful
than the existing algorithm.
Existing System:
 Existing System mainly absorbed on discovering patterns from interval events
has attracted considerable efforts due to its widespread applications.
 Sequential pattern mining is an essential data mining technique with broad
applications. Several algorithms exhibit excellent performance in discovering
sequential patterns from point-based
data.
 However, in various real-world scenarios, some events intrinsically persist for
periods of time.
 The data are typically a sequence of intervals with both start
and finish times, such as, clinical data, appliance usage data,and library
lending data, to name a few.
Disadvantages:
 The pairwise relation between two time intervals is complex .
 It may increase candidate generation and the workload for counting the
support of candidate sequence .
 Subsequently many areas of investigation includes data on a snapshot of time
points as well as time intervals, it is necessary to define a new temporal
pattern.
Proposed System:
 In this System, we progress a novel algorithm, Play to play Mining
Technique, to efficiently discover two types of interval-based sequential
patterns.
 Some Trimming techniques are proposed to further reduce the search space
of the mining process.
 Experimental studies show that proposed algorithm is efficient
and scalable. Furthermore, we apply proposed method to real
datasets to demonstrate the feasibility of discussed patterns.
Advantages:
 To reduce the search space and avoid off-putting processes.
 This proposed algorithms are both efficient and accessible, and outpace the
state-of-the-art algorithms.
References:
 R. Agrawal and R. Srikant, “Mining Sequential Patterns,” IEEE
ICDE’95, pp. 3-14, 1995.
 J. Allen, "Maintaining Knowledge about Temporal Intervals,”
Communications of ACM, vol.26, issue 11, pp.832-843, 1983.
 P. Papapetrou, G. Kollios, S. Sclaroff, and D. Gunopulos, “Discovering
frequent arrangements of temporal intervals,” IEEE ICDM’05,pp.354361,2005.
 D. Patel, W. Hsu and M. Lee, “Mining Relationships Among Intervalbased
Events for Classification,” ACM SIGMOD’08, pp. 393-404, 2008.
 J. Pei, J. Han, B. Mortazavi-Asl, H. Pinto, Q. Chen, U. Dayal, and M.C.Hsu,
“PrefixSpan: Mining Sequential Patterns Efficiently by PrefixProjected
Pattern Growth,” IEEE ICDE’01, pp. 215-224, 2001.
 E. Winarko and J.F Roddick, “ARMADA-An algorithm for discovering
 richer relative temporal association rules from interval-based data,” Data&
Knowledge Engineering, 63(1), pp. 76-90, 2007.

S. Wu and Y. Chen, “Mining Nonambiguous Temporal Patterns for IntervalBased Events,” IEEE Transactions on Knowledge and Data Engineering,
19(6), pp. 742-758,2007.