...to the project Wiki of realKD - a free open-source Java library that has been designed to help real users discovering real knowledge from real data. As such it has a strong focus on
a) a detailed data model that allows the specification of a lot of domain-dependent semantics,
b) a generic pattern model intended to capture meaningful bits of knowledge, and
c) a model to describe algorithms and their parameters in a way that makes it easy to build user-friendly applications on top of it.
In particular realKD can be used to discover associations (itemset patterns), exceptional model patterns (subgroups), and subspace outliers from tabular data. While the main purpose of realKD is to be used by other Java applications that aim to enable knowledge discovery functionality to their users, it can also be used stand-alone from the command line.
The realKD library is published via this web site under the MIT License. If you want to
- use realKD in your Java project managed by Maven look here for the dependency declaration and a short example code,
- prepare a data file to analyze with a realKD algorithm then check out the XARF file format and how to run a job from the command line,
- report an issues or missing feature you can create a card in our realKD Jira project,
- develop realKD or your own application using realKD then check out the realKD architecture.
Most of the time, it is best to also get in touch with us directly in order to discuss your ideas.
We appreciate your interest.
The realKD lead authors,