Learning In a three-Valued Environment

LIVE (Learning In a three-Valued Environment) is a system that learns in a three-valued setting: it learns both a definition for a concept and its opposite. In order to represent both positive and negative information, the representation language adopted is the one Extended Logic Programs (ELP) under the well-founded semantics extended with explicit negation (WFSX, [1]).   The WFSX semantics allows us to exploit the SLX top-down proof procedure for ELP [1]. See the paper [2] for more details on the system.

LIVE is written in XSB prolog and employs the GOLEM system. In order to run LIVE, both XSB and GOLEM must be installed in the system (you can download the binary code of GOLEM compiled for Solaris from here). A command golem must be present in the path. Due to the availability of GOLEM executables only for Solaris, at the moment this is the platform on which LIVE can be run. LIVE is composed of three XSB prolog files:

plus the code for the SLX procedure slx2.P

The input to LIVE is composed by two files: one background file and a training set file. If <file_stem> is the name given to the learning experiment, the two files will have the names <file_stem>_bg.P and <file_stem>_tr.P.

In order to run LIVE, copy both the program files and the input files in the same directory, call XSB and load LIVE by consulting the file live.P with the command
| ?-[live].
Then induction can be started with the command
| ?-i(<file_stem>).
The output is written to the file <file_stem>_rules.P

Two sample learning experiments are available (taken from [2]):

Please send a mail to Fabrizio Riguzzi if you download the code. Contact him also to report comments, bugs and new experiments.


[1] J. J. Alferes and L. M. Pereira, Reasoning with Logic Programming, volume 1111 of LNAI, Springer-Verlag, 1996.

[2] E. Lamma, F. Riguzzi, L. M. Pereira, "Strategies in Combined Learning via Logic Programs", Machine Learning, Vol. 38, Number 1 and 2, January/February 2000.

[3] E. Lamma, F. Riguzzi, L. M. Pereira, "Learning with Extended Logic Programs", Proceedings of the Logic Programming track of the Seventh International Workshop on Nonmonotonic Reasoning (LP-NMR98), May1998.

[4] E. Lamma, F. Riguzzi, L. M. Pereira, "Strategies for Learning with Extended Logic Programs", Proceedings of the Fourth International Workshop on Multistrategy Learning (MSL98), June 1998.

[5] E. Lamma, F. Riguzzi, L. M. Pereira, "Learning in a Three-valued Setting", Proceedings of the CompulogNet Area Meeting on Computational Logic and Machine Learning, June 1998.

[6] E. Lamma, F. Riguzzi, L. M. Pereira, "Learning Three-valued Logic Programs", Late breaking paper, 9th International Workshop on Inductive Logic Programming (ILP99), S. Dzeroski e P. Flach (eds.), June 1999.



Mail Your Comments on these Pages to the LIA WebMaster Mail Your Comments for LIA to the LIA PostMaster

Back to the LIA Home Page Back to the LIA Who's Who Page
Go to the DEIS Staff Page

Go to the DEIS Home Page Go to the Alma Mater Home Page