Tupleware

Warning! The following program and its description is meant for people experienced in FCA ;-)

Tupleware implements a scaling approach based on relations. It works on a set of tuples which is treated as a crossproduct of the value sets of the columns. These tuples can be read from a text file, queried via SQL (as rows of a result set), created via command line calls (stdout is read in the same way as the text files) or (via plugin) queried with RDQL from RDF/N3 files.

For scaling the object and attribute sets are selected as columns. If multiple columns are selected the crossproduct is used. The binary relation used is the projection of the original tuples onto the crossproduct of the object and attribute sets. The object set is chosen once, afterwards multiple attribute sets can be selected to create a complete CSX file which can then be loaded into ToscanaJ for filtering and nesting.

The program can be downloaded from the Sourceforge download page, it requires a recent version of Java to be installed (at least JRE/JDK 1.4.0). If you want to use the RDQL plugin you have to unzip it into the plugin directory found in the normal distribution.