Ruleminer is an experimental tool for performing various data-mining tasks, especially focused on searching for association rules. The aim is to research the association rules mining process and to develop new approaches for solving the related problems.
- Multiplatform (Java)
- Open-source (GNU General Public License)
- Reading data from CSV
- Searching for impact rules while utilizing various pruning mechanisms that speed-up the search process:
- Searching top n best rules
- Pruning trivial rules
- Searching for "very interesting" impact rules
- Allowing both positive and negative predicates in the conditions
This project was partially supported by the internal grant number SGS1/PDF/2011 of the University of Ostrava.