CoStar

CO-clustering for STAR-structured data (CoStar) is a multi-view co-clustering algorithm. Given a set of different views of the same dataset CoStar provides, at the same time, a clustering on documents and a clustering on features of each view.

If using the code, please cite the following paper:

[1] Ienco, Dino, et al., 2013. Parameter-less co-clustering for star-structured heterogeneous data. Data Mining and Knowledge Discovery 26.2: 217-254

Run the example

• Compile the sources

  make all

• Enter the example data folder

  cd execution-example

• Run the program with the example data contained in the folder

  ../multiViewCoClu -r 6 -n 100 -i config.txt -k 2 -o output

Parameters of the executable

The executable produced after sources compilation takes in input the following parameters:

• -r <int>, the number of rows (documents) contained in the dataset
• -n <int>, the number of costar iterations to perform. Consider that each iteration corresponds to a single element move.
• -i <filepath>, the path (or simply the name) of the configuration file of the current dataset
• -o <string>, prefix for the output files
• -k <int>, the minimum number of cluster to give as output

Dataset files

The dataset should be composed by:

• a txt file for each considered view, these files contains a row for each document in the dataset. Each row is a space separated sequence of values
• a configuration file, that contains a row for each view in the dataset. Each row contains two space separated values: <n_features, view_file>, where
  • n_features is the number of features of the considered view, and
  • view_file is the name of the file containing values for the view.

Output files

The program produces the following files, where 'prefix' is the prefix specified from the command line.

• prefix.row - contains the row clusters assignment
• prefix_N.col - N files with N from 0 to n_views. Each file contains the column clusters assignment for the i-th view.
• prefix.time - contains the execution time
• prefix.iter - contains for each iteration the tau_r and tau_c values

Contributions

Original implementation by Dino Ienco, refactoring by Valentina Rho.

Random Number generator library of type 'Mersenne Twister', by Agner Fog (http://www.agner.org/random/) described in:

M. Matsumoto & T. Nishimura. Mersenne Twister: A 623-Dimensionally Equidistributed Uniform Pseudo-Random Number Generator. ACM Transactions on Modeling and Computer Simulation, Vol. 8, no. 1, 1998, pp. 3-30.

Licence

GNU General Public License v3.

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

For details of the GNU General Public License refer to http://www.gnu.org/licenses/.