diff --git a/packages/cplint/doc/bib.bib b/packages/cplint/doc/bib.bib index 50460c5c8..c9e6f9193 100644 --- a/packages/cplint/doc/bib.bib +++ b/packages/cplint/doc/bib.bib @@ -1,3 +1,14 @@ +@article{DBLP:journals/ai/Cohen95, + author = {William W. Cohen}, + title = {Pac-Learning Non-Recursive Prolog Clauses}, + journal = {Artif. Intell.}, + volume = {79}, + number = {1}, + year = {1995}, + pages = {1-38}, + ee = {http://dx.doi.org/10.1016/0004-3702(94)00034-4}, + bibsource = {DBLP, http://dblp.uni-trier.de} +} @article{BelRig13-TPLP-IJ, author = {Elena Bellodi and Fabrizio Riguzzi}, title = {Structure Learning of Probabilistic Logic Programs by Searching the Clause Space}, diff --git a/packages/cplint/doc/manual.html b/packages/cplint/doc/manual.html index 566009765..5bbf13c0e 100644 --- a/packages/cplint/doc/manual.html +++ b/packages/cplint/doc/manual.html @@ -7,7 +7,7 @@ - + @@ -27,12 +27,12 @@ class="cmr-12">September 4, 2013 id="x1-10001">Introduction

cplint is a suite of programs for reasoning with ICL [14], LPADs [23, 24] and +href="#XDBLP:journals/ai/Poole97">15], LPADs [24, 25] and CP-logic programs [21, 22]. It contains programs both for inference and +href="#XVenDenBru-JELIA06">22, 23]. It contains programs both for inference and learning.

2

The coin example of [24] is represented as (see file 25] is represented as (see file coin.cpl)

@@ -169,13 +169,13 @@ class="cmtt-10">cplint contains various modules for answering queries. +href="#XRig11-CILC11-NC">19]
  • approx/exact.pl as P, i.e., sufficient causation, independent causation, no deus ex machina events and temporal precedence. It uses the definition of the semantics given in [22].
    • +href="#XDBLP:journals/tplp/VennekensDB09">23].

    4.1 Commands

    @@ -827,44 +827,44 @@ class="cmtt-10">examples:
  • alarm.cpl: representation of the Bayesian network in Figure 2 of [24]. +href="#XVenVer04-ICLP04-IC">25].
  • coin.cpl: coin example from [24]. +href="#XVenVer04-ICLP04-IC">25].
  • coin2.cpl: coin example with two coins.
  • dice.cpl: dice example from [24]. +href="#XVenVer04-ICLP04-IC">25].
  • twosideddice.cpl, threesideddice.cpl game with idealized dice with two or three sides. Used in the experiments in [16]. +href="#XRig-RCRA07-IC">17].
  • ex.cpl: first example in [16]. +href="#XRig-RCRA07-IC">17].
  • exapprox.cpl: example showing the problems of approximate inference (see [16]). +href="#XRig-RCRA07-IC">17]).
  • exrange.cpl: example showing the problems with non range restricted programs (see [16]). +href="#XRig-RCRA07-IC">17]).
  • female.cpl: example showing the dependence of probabilities in the head from variables in the body (from [24]). +href="#XVenVer04-ICLP04-IC">25]).
  • mendel.cpl, mendels.cpl: programs describing the Mendelian @@ -875,7 +875,7 @@ href="#XBlo04-ILP04WIP-IC">7]. class="cmtt-10">paper_ref.cpl, paper_ref_simple.cpl: paper citations examples, showing reference uncertainty, inspired by [13]. +href="#XGetoor+al:JMLR02">14].
  • paper_ref_not.cpl: paper citations example showing that negation @@ -895,13 +895,13 @@ class="cmtt-10">school.cpl.
  • student.cpl: student example from Figure 1.3 of [12]. +href="#XGetFri01-BC">13].
  • win.cpl, light.cpl, trigger.cpl, throws.cpl, hiv.cpl,
    invalid.cpl: programs taken from [22]. 23]. invalid.cpl is an example of a program that is invalid but sound.

    • The files cplint contains the following learning algorithms: class="cmtt-10">cplint EM): an implementation of EM for learning parameters that is based on lpadsld.pl [19] +href="#XRigDiM11-ML-IJ">20]

  • RIB (Relational Information Bottleneck): an algorithm for learning parameters based on the Information Bottleneck [19] +href="#XRigDiM11-ML-IJ">20]
  • EMBLEM (EM over Bdds for probabilistic Logic programs Efficient Mining): an implementation of EM for learning parameters that computes @@ -1253,7 +1253,7 @@ class="cmtt-10">logsize_fraction times its maximum class="cmsy-10">|CH,T|, see [11]) +href="#XDBLP:journals/jmlr/ElidanF05">12])
  • delta (values: negative integer, default value -10, valid for RIB): value @@ -1263,7 +1263,7 @@ class="cmtt-10">delta (values: negative integer, default va class="cmtt-10">epsilon_fraction (values: integer, default value 100, valid for RIB): in the computation of the step, the value of ϵ of [11] is obtained as +href="#XDBLP:journals/jmlr/ElidanF05">12] is obtained as log |CH,Tmax_iter_structure (values: integer, d
  • background_clauses (values: integer, default value: 50, valid for - SLIPCOVER): maximum numbers of background clauses.
    • -

    + SLIPCOVER): maximum numbers of background clauses + +

  • maxdepth_var (values: integer, default value: 2, valid for SLIPCOVER): + maximum depth of variables in clauses (as defined in [10]).
    • +

    5.3 Commands

    -

    To execute CEM, load

    To execute CEM, load em.pl with

    ?:- use_module(library(’cplint/em’)).
    -

    and call: +

    and call:

    ?:- em(stem).
    -

    To execute RIB, load

    To execute RIB, load rib.pl with

    ?:- use_module(library(’cplint/rib’)).
    -

    and call: +

    and call:

    ?:- ib_par(stem).
    -

    To execute EMBLEM, load

    To execute EMBLEM, load slipcase.pl with

    ?:- use_module(library(’cplint/slipcase’)).
    -

    and call +

    and call

    ?:- em(stem).
    -

    To execute SLIPCASE, load

    To execute SLIPCASE, load slipcase.pl with

    ?:- use_module(library(’cplint/slipcase’)).
    -

    and call +

    and call

    ?:- sl(stem).
    -

    To execute SLIPCOVER, load

    To execute SLIPCOVER, load slipcover.pl with

    ?:- use_module(library(’cplint/slipcover’)).
    -

    and call +

    and call

    ?:- sl(stem).
    -

    +

    5.4 Learning Examples

    -

    The subfolders

    The subfolders em, rib, slipcase and slipcover of the class="cmtt-10">packages/cplint folder in Yap git distribution contain examples of input and output files for the learning algorithms. -

    +

    6 License

    -

    cplint, as Yap, follows the Artistic License 2.0 that you can find in Yap CVS root dir. The copyright is by Fabrizio Riguzzi. -

    The modules in the approx subdirectory use SimplecuddLPADs, a modification of +

    The modules in the approx subdirectory use SimplecuddLPADs, a modification of the Simplecudd library whose copyright is by Katholieke Universiteit Leuven and that follows the Artistic License 2.0. -

    Some modules use the library

    Some modules use the library CUDD for manipulating BDDs that is included in glu. For the use of CUDD, the following license must be accepted: -

    Copyright (c) 1995-2004, Regents of the University of Colorado -

    All rights reserved. -

    Redistribution and use in source and binary forms, with or without modification, +

    Copyright (c) 1995-2004, Regents of the University of Colorado +

    All rights reserved. +

    Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

    -

    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS

    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS
    AND CONTRIBUTORS ”AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR @@ -1447,7 +1452,7 @@ class="newline" />AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -

    lpad.pl, semlpad.pl and cpl.pl are based on the SLG system by Weidong @@ -1534,29 +1539,36 @@ class="cmti-10">Journal of the ACM, 43(1):20–74, 1996.

    [10]   William W. Cohen. Pac-learning non-recursive prolog clauses. Artif. + Intell., 79(1):1–38, 1995. +

    +

    + [11]   L. De Raedt, A. Kimmig, and H. Toivonen. ProbLog: A probabilistic Prolog and its application in link discovery. In International Joint Conference on Artificial Intelligence, pages 2462–2467, 2007. -

    -

    - [11]   G. Elidan and N. Friedman. Learning hidden variable networks: The - information bottleneck approach. Journal of Machine Learning Research, - 6:81–127, 2005.

    [12]   G. Elidan and N. Friedman. Learning hidden variable networks: The + information bottleneck approach. Journal of Machine Learning Research, + 6:81–127, 2005. +

    +

    + [13]   L. Getoor, N. Friedman, D. Koller, and A. Pfeffer. Learning probabilistic relational models. In Saso Dzeroski and Nada Lavrac, editors, Relational Data Mining. Springer-Verlag, Berlin, 2001.

    - [13]      L. Getoor, N. Friedman, D. Koller, and B. Taskar. Learning probabilistic models of relational structure. Journal of Machine Learning @@ -1564,13 +1576,13 @@ class="cmti-10">Journal of Machine Learning class="cmti-10">Research, 3:679–707, December 2002.

    - [14]      David Poole. The independent choice logic for modelling multiple agents under uncertainty. Artificial Intelligence, 94(1-2):7–56, 1997.

    - [15]      Fabrizio Riguzzi. A top down interpreter for LPAD and CP-logic. In Congress of the Italian Association for Artificial Intelligence, volume 4733 @@ -1578,7 +1590,7 @@ class="cmti-10">Congress of the Italian Association for Artificial Intelligence< class="cmti-10">LNAI, pages 109–120. Springer, 2007.

    - [16]      Fabrizio Riguzzi. A top down interpreter for LPAD and CP-logic. In Proceedings of the 14th RCRA workshop Experimental Evaluation of @@ -1586,13 +1598,13 @@ class="cmti-10">Proceedings of the 14th RCRA workshop Experimental Evaluation of class="cmti-10">Algorithms for Solving Problems with Combinatorial Explosion, 2007.

    - [17]      Fabrizio Riguzzi. Extended semantics and inference for the Independent Choice Logic. Logic Journal of the IGPL, 17(6):589–629, 2009.

    - [18]      Fabrizio Riguzzi. MCINTYRE: A Monte Carlo algorithm for probabilistic logic programming. In Proceedings of the 26th Italian @@ -1600,17 +1612,17 @@ class="cmti-10">Proceedings of the 26th Italian class="cmti-10">Conference on Computational Logic (CILC2011), Pescara, Italy, 31 August-2 September, 2011, 2011. -

    -

    - [19]   Fabrizio Riguzzi and Nicola Di Mauro. Applying the information - bottleneck to statistical relational learning. Machine Learning, 2011. To - appear.

    [20]   Fabrizio Riguzzi and Nicola Di Mauro. Applying the information + bottleneck to statistical relational learning. Machine Learning, 2011. To + appear. +

    +

    + [21]   V. Santos Costa, D. Page, M. Qazi, and J. Cussens. CLP(BUncertainty class="cmti-10">in Artificial Intelligence. Morgan Kaufmann, 2003.

    - [21]      J. Vennekens, M. Denecker, and M. Bruynooghe. Representing causal information about a probabilistic process. In Proceedings of the 10th @@ -1630,19 +1642,19 @@ class="cmti-10">European Conference on Logics in Artificial Intelligence, September 2006.

    - [22]      J. Vennekens, Marc Denecker, and Maurice Bruynooghe. CP-logic: A language of causal probabilistic events and its relation to logic programming. Theory Pract. Log. Program., 9(3):245–308, 2009.

    - [23]      J. Vennekens and S. Verbaeten. Logic programs with annotated disjunctions. Technical Report CW386, K. U. Leuven, 2003.

    - [24]      J. Vennekens, S. Verbaeten, and M. Bruynooghe. Logic programs with annotated disjunctions. In International Conference on Logic diff --git a/packages/cplint/doc/manual.pdf b/packages/cplint/doc/manual.pdf index 401cac50c..216e2ef4d 100644 Binary files a/packages/cplint/doc/manual.pdf and b/packages/cplint/doc/manual.pdf differ diff --git a/packages/cplint/doc/manual.tex b/packages/cplint/doc/manual.tex index 755cf20b6..bcf914e76 100644 --- a/packages/cplint/doc/manual.tex +++ b/packages/cplint/doc/manual.tex @@ -626,7 +626,9 @@ If set to \verb|atoms|, a BDD is built for the conjunction of a group of atoms f \item \verb|max_iter_structure| (values: integer, default value: 1, valid for SLIPCOVER): maximum number of theory search iterations \item \verb|background_clauses| (values: integer, default value: 50, valid for SLIPCOVER): - maximum numbers of background clauses. + maximum numbers of background clauses +\item \verb|maxdepth_var| (values: integer, default value: 2, valid for SLIPCOVER): maximum depth of +variables in clauses (as defined in \cite{DBLP:journals/ai/Cohen95}). \end{itemize} \subsection{Commands} To execute CEM, load \texttt{em.pl} with