------------------------------------------ General Information ----------------------------------------

This section contains the general information about our code and dataset. This folder contains our proposed algorithm, SGPursuit, which is under review of ICDM-2017.

In this folder, there are three sub-folders:

1. code: contains our SGPursuit algorithm and the baseline methods used in our paper.
2. dataset: has 3 datasets that we used in our experiments.
3. dataProcessing: contains the code and data for preprocessing of the real data.

Firstly, go to {your/root/path}/code/SGPursuit/ directory. There are 2 ways to run SGPursuit.

1. use mvn exec:java command: mvn exec:java -Dexec.mainClass="edu.albany.cs.SGPursuit.TestOnCrimesOfChicago" for Crimes of chicago mvn exec:java -Dexec.mainClass="edu.albany.cs.SGPursuit.TestOnYelp" for Yelp data mvn exec:java -Dexec.mainClass="edu.albany.cs.DenseSubGraph.DenseSubGraphUnitTest" for simulation data mvn exec:java -Dexec.mainClass="edu.albany.cs.DenseSubGraph.MultiClusterSGPursuit.RealdataSetsExperiment" for Real dataset data !!!Important: You should install maven and Java-1.8 before you run this program.

[Datasets]: [Simualtion Dataset Main Folder] data/DenseGraph/DenseSubgraph_APDM/ [Real world Dataset(GraphML files and apdm files) ] data/DenseGraph/Realdata/apdm/

2. use Intellij IDE: You can use Intellij IDE to import our code as a maven project. After you set up the project, then you can find TestOnCrimesOfChicago.java, TestOnYelp.java, and DenseSubGraphUnitTest.java. Finally, you can run it.

Baseline methods are FocusCO,GAMer,SODA,AMEN. SODA and AMEN are in code main folder. FocusCO and GAMer are in "{SGPursuit project}/edu/albany/cs/baselines/".

1. AMEN: go to the folder {your/root/path}/code/AMEN. Make sure you have matlab environment. Then just type the following command to go to maltab environment: >>matlab -nodesktop >>test_amen_on_crimes_of_chicago
2. SODA: go to {your/root/path}/code/SODA. Make sure you have python and cvxopt (and solvers). >>python soda_crimesOfchicago.py
3. FocusCO: [Simualtion Dataset Main Folder] /SG-Pursuit/data/DenseGraph/DenseSubgraph_APDM/ [Sub foldors] 1)VaryingNumOfAttributes 2)VaryingNumOfClusters 3)VaryingClusterSizes
   Requirements: 1. A recent version of Matlab 2.Java 6+ There are two programs in this baseline method. The first is a matlab script which learns a distance metric and reweighs the input graph. The second is a java program which extracts communities & outliers from the reweighted graph. We use a matlab proxy tool to call matlab code from the java project.

To run FocusCO use maven: "mvn exec:java -Dexec.mainClass="edu.albany.cs.baselines.FocusCO.FocusCOMatlab.TestFocusCO.focusCOTest1VaryNumOfAtt" this expeiment function coressponding to the Figure 4 in our paper.(or you can test other experiment function in same class: focusCOTest1VaryNumOfAtt(),focusCOTest2VaryNumOfCluster(), focusCOTest3VaryClusterSize(). Note*: you can download the original code from github: https://github.com/phanein/focused-clustering.

4. GAMer: This method provided by the authors in page: "http://dme.rwth-aachen.de/en/gamer". They provided a executable jar file. You can find our test code in "edu.albany.cs.baselines.GAMer". 4.1 Simualtion dataset experiment [Simualtion Dataset Main Folder] /SG-Pursuit/data/DenseGraph/DenseSubgraph_APDM/ [Sub foldors] 1)VaryingNumOfAttributes 2)VaryingNumOfClusters 3)VaryingClusterSizes To run GAMer test code use maven: "mvn exec:java -Dexec.mainClass="edu.albany.cs.baselines.GAMer.TestGAMer.test1VaryNumOfAtt" this expeiment resutls coressponding to the Figure 4 in our paper.(or you can test other experiment function in same class: test2VaryNumOfCluster(),test3VaryClusterSize() 4.2 Real dataset experiemnt [Real world Dataset(GraphML files) ] /SG-Pursuit/data/DenseGraph/Realdata/. this expeiment resutls coressponding to the Table 4 in our paper. To run GAMer realdata test code use maven: "mvn exec:java -Dexec.mainClass="edu.albany.cs.baselines.GAMer.TestGAMerRealDataset.testRealdatasets"

In case that you are interested in our data preprocessing. The folder dataProcessing contains the raw data and the pre-processing code. The pre-processing code is not documented very well. However, you can email me (bzhou6@albany.edu / aalimu@albany.edu) if you have any questions.