Database Reference
In-Depth Information
Chapter 2
Frequent Pattern Mining Algorithms: A Survey
Charu C. Aggarwal, Mansurul A. Bhuiyan and Mohammad Al Hasan
Abstract
This chapter will provide a detailed survey of frequent pattern mining
algorithms. A wide variety of algorithms will be covered starting from
Apriori
.
Many algorithms such as
Eclat
,
TreeProjection
, and
FP-growth
will be discussed.
In addition a discussion of several maximal and closed frequent pattern mining
algorithms will be provided. Thus, this chapter will provide one of most detailed
surveys of frequent pattern mining algorithms available in the literature.
·
·
·
Keywords
Frequent
pattern
mining
algorithms
Apriori
TreeProjection
FP-growth
1
Introduction
In data mining, frequent pattern mining (FPM) is one of the most intensively inves-
tigated problems in terms of computational and algorithmic development. Over the
last two decades, numerous algorithms have been proposed to solve frequent pattern
mining or some of its variants, and the interest in this problem still persists [
45
,
75
].
Different frameworks have been defined for frequent pattern mining. The most com-
mon one is the support-based framework, in which itemsets with frequency above
a given threshold are found. However, such itemsets may sometimes not represent
interesting positive
correlations
between items because they do not normalize for
the absolute frequencies of the items. Consequently, alternative measures for inter-
estingness have been defined in the literature [
7
,
11
,
16
,
63
]. This chapter will focus
on the support-based framework because the algorithms based on the interestingness
·
