Knowledge Discovery Group at FI MU Brno

Knowledge Discovery Group at FI MU Brno

Projects


DIODA

""""DIODA, an efficient implementation of secure mining of association rules in horizontally partitioned data, is one of the first systems for secure distributed mining of association rules.

Author: Jan Frieser .::. Contact: Lubomir Popelinsky (popel) .::. State: finished .::. URL: http://www.fi.muni.cz/kd/projects/dioda/kriticka_zprava.pdf

e-learning and data mining (in Czech)

""Experiments with data mining methods for analysis of e-learning process, which is a part of IS MU, are described.

Author: Martin Vecera and Petra Turkova .::. Contact: Lubomir Popelinsky (popel) .::. State: finished .::. URL: http://www.fi.muni.cz/kd/projects/e-learning_DM/zprava.pdf

FactBuilder

FactBuilder is a tool for transforming propositional data into format appropriate for ILP systems.

Author: Petra Turková .::. Contact: Petra Turková (xturkova) .::. State: running .::. URL: http://www.fi.muni.cz/kd/projects/factbuilder/

GRR

GRR integrates GRASS 5 with statistical tool R, Weka and some other data mining tools.

Author: (Tomáš Buk, Patrik Steciv), Tomáš Ludík, Martin Wokoun .::. Contact: Jan Blaťák (xblatak) .::. State: running .::. URL: http://www.fi.muni.cz/kd/projects/grr/

ILPVisualizer

ILPVisualizer is a postprocessing tool for Inductive Logic Programming (ILP) which displays first-order formulae as planar graphs.

Author: Jindřich Březina .::. Contact: Jindřich Březina .::. State: finished .::. URL: http://www.fi.muni.cz/kd/projects/ilpvisualizer/

Indeed

Indeed is a general-purpose inductive logic programming system which is aimed at efficient introduction of constants into the body of a clause. These constants can be arbitrary terms, not only atoms.

Author: Miloslav Nepil .::. Contact: Jan Blaťák (xblatak) .::. State: finished

OR-FP

OR-FP is an algorithm for mining frequent patterns in object-oriented and object-relatinal data. The OR-FP system was implemented in Java and it works with data stored in the Oracle object-relational database system.

Author: Petr Kuba .::. Contact: Jan Blaťák (xblatak) .::. State: finished

RAP

RAP is the first system for searching first-order maximal frequent patterns efficiently. Three search strategies have been implemented in RAP: random, depth-first search and best-first search. Different pre-defined as well as user-defined heuristics can be employed in RAP.

Author: Jan Blaťák (xblatak) .::. Contact: Jan Blaťák (xblatak) .::. State: running .::. URL: http://www.fi.muni.cz/kd/projects/rap/

dRAP

dRAP is an extension of the RAP system. It is intended for distributed mining. Now the dRAP-Independent algorithm which is able to mine patterns in horizontally fragmented data is implemented.

Author: Jan Blaťák (xblatak) .::. Contact: Jan Blaťák (xblatak) .::. State: running

Animations for teaching logic

"""Animations for computational logic (interpretations, truth tables, resolution, tableaux proofs, inductive inference) Authors: Jan Vaňáček Zdeněk Zavrtal Marek Stehlík Eva Mráková Luboš Popelínský

Author: KD lab .::. Contact: popel@fi.muni.cz .::. State: running .::. URL: http://www.fi.muni.cz/kd/projects/animationsForLogic

KD Group © 2008