swMATH ID: 10969
Software Authors: Øhrn, Aleksander; Komorowski, Jan; Skowron, Andrzej; Synak, Piotr
Description: The design and implementation of a knowledge discovery toolkit based on rough set – the ROSETTA system. The KDD process using rough sets has been presented and analyzed. Following the requirement specifications of a sophisticated user-environment for empirical model construction, the design and implementation of a software toolkit has been outlined. The resulting toolkit covers the whole range of KDD tasks within the realm of rough sets. It consists of a general C++ class library primarily aimed at researchers for rapid prototyping, and a GUI front-end developed for knowledge discovery in an interactive setting. Issues springing from the overall KDD process have been the principal guiding design parameters for both.par The kernel class library provides a set of fundamental building blocks and the means to combine these in a flexible fashion, both for the development and testing of new algorithms and for partial automation of the overall KDD process. Various design choices made during construction of the class library have been outlined, and examples of its use been given. The GUI offers an environment wherein the fundamental tools furnished by the kernel are set. This enables interactive manipulation and creation of objects related to the KDD process. Jointly, the kernel and the front-end offer a means to effectively and easily conduct KD and data mining experiments within the framework of rough set theory.par In a few aspects, the work on ROSETTA is related to our previous research on providing tools and programming environments for process-oriented synthesis of logic programs.
Homepage: http://www.lcb.uu.se/tools/rosetta/resources.php
Keywords: KDD process; rough sets
Related Software: UCI-ml; LERS; RSES; Rseslib; WEKA; C4.5; RIONA; jMAF; RoughSets; clusfind; ReproZip; Vertica; DAGGER; RSL; KDD Cup; FSMRDE; CMAR; FOIL; 4eMka2; DIXER
Cited in: 20 Publications

Citations by Year