ca.pfv.spmf.algorithms.sequentialpatterns.fournier2008_seqdim.multidimensionalpatterns

Class AlgoDim

java.lang.Object

ca.pfv.spmf.algorithms.sequentialpatterns.fournier2008_seqdim.multidimensionalpatterns.AlgoDim

public class AlgoDim
extends java.lang.Object

Implementation of the DIM algorithm by Pinto et al. (2001) to extract frequent MD-Patterns (multi-dimensional patterns) from a MD-Database. The algorithm is described in:

Pinto, H., Han, J., Pei, J., Wang, K., Chen, Q., & Dayal, U. (2001, October). Multi-dimensional sequential pattern mining. In Proceedings of the tenth international conference on Information and knowledge management (pp. 81-88). ACM.

This implementation use the Apriori, AprioriClose or CHARM algorithms depending on what the user prefers. This allow to find all frequent MD-Patterns or just those that are closed.

The idea of closed MD sequential pattern mining is described in (Songram, 2006):

P. Songram, V. Boonjing, S. Intakosum: Closed Multi-dimensional Sequential-Pattern Minin. Proc. of ITNG 2006.

This algorithm implementation proceeds as follow, it (1) convert MD-Patterns into itemsets, (2) mine frequent (closed) itemsets from the md-patterns generated in Step 1 and (3) convert frequent (closed) itemsets back in MD-Patterns.

See Also:

MDPattern, MDPatterns, MDPatternsDatabase

Constructor Summary

Constructors

Constructor and Description
AlgoDim(boolean findClosedPatterns, boolean findClosedPatternsWithCharm)

Method Summary

Modifier and Type	Method and Description
void	printStats(int databaseSize) Print statistics about this algorithm execution to System.out
MDPatterns	runAlgorithm(MDPatternsDatabase mdPatDatabase, double minsupp) Run the DIM algorithm

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

AlgoDim

public AlgoDim(boolean findClosedPatterns,
               boolean findClosedPatternsWithCharm)

Parameters:

findClosedPatterns - Indicates if this class has to find respectively frequent itemsets or frequent closed itemsets.

findClosedPatternsWithCharm - if true, the algorithm finds closed patterns with Charm instead // of AprioriClose

Method Detail

runAlgorithm

public MDPatterns runAlgorithm(MDPatternsDatabase mdPatDatabase,
                               double minsupp)
                        throws java.io.IOException

Run the DIM algorithm

Parameters:

mdPatDatabase - an md-pattern database

minsupp - a minimum support threshold as a percentage (double)

Returns:

the md-patterns found

Throws:

java.io.IOException - exception if error reading/writing file

printStats

public void printStats(int databaseSize)

Print statistics about this algorithm execution to System.out

Parameters:

databaseSize - the number of mdpattern in the md-pattern database.