Pattern Space Maintainer (PSM)
Version 2.0
==============================

PSM is a software for maintaining the frequent pattern space (in terms of 
equivalent classes) when insertions/deletions are made to the transaction
database. It also handles changes in the pattern space due to change in
the support threshold.



Usage of PSM

Datasets:
t8_2.dat --- orgininal dataset
t8_2.in	--- incremental dataset

t8_20p.close --- existing equivelance classes information
t8_full_20p.tree --- intermediate file
t8_full_20p.neg --- intermediate file
t9_full_20p.key --- existing generators

1. One needs to generate the existing frequent patterns using ECgen.

usage: ecgen <infile> <outfile> <min sup>
example: ecgen t8_2.dat t8_full_20p 0.2

2. One now can maintain the frequent patterns with PSM+

usage: psm [options...]
-e existing pattern file name
-i incremental file name
-d decremental file name
-iu incremental update size
-du decremental update size
-s min support threshold
-o output file name
-p print or not

example for incremental update:

psm -e t8_full_20p -i t8_2.in -iu 2 -s 0.2 -o test

example for support adjustment:

psm -e t8_full_20p -s 0.1 -o test

example for decrementla update:

psm -e t8_full_20p -d t8_2.dec -du 2 -s 0.2 -o test


Credits:

This software was developed by Mengling Feng as part of his PhD thesis.
The work was supported by a A*STAR AGS scholarship and by MOE Tier-1 grant
R-252-040-238-101/R-252-060-238-133 and A*STAR SERC PSF grant
SERC 072 101 0016.


If you use this software, please cite the following papers:

Mengling Feng, Jinyan Li, Limsoon Wong, Yap-Peng Tan. 
Negative Generator Border for Effective Pattern Maintenance. 
Proceedings of 4th International Conference on Advanced Data Mining and Applications, 
pages 217--228, Chengdu, China, 8-10 October 2008.