Togaware: The Gnome Data Mine

The Gnome Data Mine

The following open source tools are provided under the GNU General Public License as free open source software in the hope that others will find them useful and may even improve upon them. As the GNU General Public License states there is no warranty, nor even the implied warranty of merchantability or fitness for a particular purpose. We provide these tools in good faith for the benefit of the community.

Gnome Data Mine Tools

The gnome-datamine-tools is a growing collection of tools packaged to provide a freely available single collection of data mining tools. It is available as gnome-datamine-tools.tar.gz. This contains everything required, including the GUIs and the data mining applications. You will require Python and Gnome installed on your system, and I recommend the Debian GNU/Linux distribution. The Gnome Data Mining Tools package is simply extracted, compiled, and installed into /usr/local with:

  $ tar zxvf gnome-datamine-tools.tar.gz
  $ cd gnome-datamine-tools-<version>
  $ make install

The applications can then be run. Try gdmapriori for starters:

  $ gdmapriori

Visit the individual pages for the tools below for examples of using the tools out of the box.

The default install with make install will place the applications in /usr/local by default. To install elsewhere use make PREFIX=/usr install, for example.

The following tools are included in the package:

Apriori Association Rules

The gdmapriori application is a Gnome utility to extract association rules from transaction data. Numerous options are available. The package includes both the Gnome GUI and the apriori command line from Christian Borgelt.

Bayes Classifier

The gdmbayes application is a Gnome utility to build Bayes classifiers from training data. The package includes both the Gnome GUI and the bayes command line from Christian Borgelt.

Decision Trees

The gdmdtree application is a Gnome utility to build a decision tree from training data. Numerous options are available, including generating rules. The package includes both the Gnome GUI and the dtree command line from Christian Borgelt.

CSV Tool

The csv tool provides plotting (to xfig) and table generateion (for LaTeX). Two command line tools are provided. csv2plot will plot an x/y graph of data. csv2latex will generate the LaTeX table commands for inclusion in a LaTeX document.

Download

The Gnome Data Mining tools can be downloaded from the official project pages at Sarovar.org or directly here as gnome-datamine-tools.tar.gz.

The following tools will be migrated into the gnome-datamine-tools pacakge in the future.

Bar Chart

Bar Chart can be used to generate bar charts from supplied data files. Output is in PDF, PNG, FIG, or EPS.

Bin Chart

Bin Chart can be used to generate bar charts to display the frequency distribution of a set of data contained in a data file. The data is binned into evenly distributed (along the number line from the data's minimum value to maximum value) bins. Output is in PDF, PNG, FIG, or EPS.

GDM Plot

GDM Plot is a Gnome utility to produce binned plots in various formats, including LaTeX, PostScript, and PDF. Plots are generated from loaded data files and once the plot looks correct you can generate the plot in one of these formats.