==========================================================

    MTP - Minus Top-k Partition
    Author: Yongsub Lim ([email protected]), and U Kang
    Version: 2.0
    Release date: Dec 8, 2014
    
==========================================================

This software is free of charge under research purposes.
For commercial purposes, please contact the author.

MTP is an algorithm to find a set of nodes in a graph so that its subgraph has 
higher quality graph partitions that the original graph.

gbMTP, an extension of MTP, is an algorithm to find a global balanced partition.

For details of the methods, we refer to the following paper:
Yongsub Lim, Won-Jo Lee, Ho-Jin Choi, and U Kang. "MTP: Discovering high quality partitions in real world graphs." World Wide Web Journal, 2016.



HOW TO RUN
==
Below, we use #MTP_ROOT#, #METIS_ROOT#, and #METISMEX_ROOT# to denote
the root directories of MTP, METIS, and metismex, respectively.

1. Download Packages
MTP requries 
	a) METIS, a graph partitioning library provided in 
		http://glaros.dtc.umn.edu/gkhome/metis/metis/download, and
	b) its MATLAB interface 'metismex' provided in 
		https://github.com/dgleich/metismex.

2. Prepare METIS
Move #METIS_ROOT# below #MTP_ROOT#. 
To compile METIS, read BUILD.txt and Install.txt in #METIS_ROOT#. 
	
3. Prepare metismex 
	a) Rename #METISMEX_ROOT# to 'metismex'.
	b) Copy #METISMEX_ROOT# below #METIS_ROOT#.
	d) Run make.m in #METISMEX_ROOT#.

4. Prepare mex for MTP
Run mex_compile.m in #MTP_ROOT$.
	
4. Run demo
Run demo.m in the MATLAB command.
Our tests were done on 64bit Linux with METIS version 5.1.0.



For further questions, contact Yongsub Lim ([email protected]).