18553a6ff8
git-svn-id: https://yap.svn.sf.net/svnroot/yap/trunk@1988 b08c6af1-5177-4d33-ba66-4b1c6b8b522a
894 lines
40 KiB
HTML
894 lines
40 KiB
HTML
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
|
||
"http://www.w3.org/TR/html4/loose.dtd">
|
||
<html >
|
||
<head><title>cplint Version 1.0 Manual</title>
|
||
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
|
||
<meta name="generator" content="TeX4ht (http://www.cse.ohio-state.edu/~gurari/TeX4ht/)">
|
||
<meta name="originator" content="TeX4ht (http://www.cse.ohio-state.edu/~gurari/TeX4ht/)">
|
||
<!-- html -->
|
||
<meta name="src" content="manual.tex">
|
||
<meta name="date" content="2007-11-08 18:55:00">
|
||
<link rel="stylesheet" type="text/css" href="manual.css">
|
||
</head><body
|
||
>
|
||
<div class="maketitle">
|
||
|
||
|
||
|
||
<h2 class="titleHead">cplint Version 1.0 Manual</h2>
|
||
<div class="author" ><span
|
||
class="cmr-12">Fabrizio Riguzzi</span>
|
||
<br /><span
|
||
class="cmr-12">fabrizio.riguzzi@unife.it</span></div>
|
||
<br />
|
||
<div class="date" ><span
|
||
class="cmr-12">November 8, 2007</span></div>
|
||
</div>
|
||
<h3 class="sectionHead"><span class="titlemark">1 </span> <a
|
||
id="x1-10001"></a>Introduction</h3>
|
||
<!--l. 30--><p class="noindent" ><span
|
||
class="cmtt-10">cplint </span>is an interpreter for LPADs <span class="cite">[<a
|
||
href="#XVenVer03-TR">10</a>, <a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span> and CP-logic programs <span class="cite">[<a
|
||
href="#XVenDenBru-JELIA06">9</a>, <a
|
||
href="#XCP-logic-unp">12</a>]</span>. It is
|
||
described in <span class="cite">[<a
|
||
href="#XRig-AIIA07-IC">6</a>]</span> and <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span>. It is an adaptation of the interpreter for ProbLog
|
||
<span class="cite">[<a
|
||
href="#XDBLP:conf/ijcai/RaedtKT07">3</a>]</span>.
|
||
<!--l. 32--><p class="indent" > It was proved correct <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span> for range restricted acyclic programs <span class="cite">[<a
|
||
href="#XDBLP:journals/ngc/AptB91">1</a>]</span> without function
|
||
symbols.
|
||
<!--l. 34--><p class="indent" > It is also able to deal with extensions of LPADs and CP-logic: the clause bodies
|
||
can contain <span
|
||
class="cmtt-10">setof </span>and <span
|
||
class="cmtt-10">bagof</span>, the probabilities in the head may be depend on
|
||
variables in the body and it is possible to specify a uniform distribution in the head
|
||
with reference to a <span
|
||
class="cmtt-10">setof </span>or <span
|
||
class="cmtt-10">bagof </span>operator. These extended features have been
|
||
introduced in order to represent CLP(BN) <span class="cite">[<a
|
||
href="#XSanPagQaz03-UAI-IC">8</a>]</span> programs and PRM models <span class="cite">[<a
|
||
href="#XGetoor+al:JMLR02">5</a>]</span>: <span
|
||
class="cmtt-10">setof</span>
|
||
and <span
|
||
class="cmtt-10">bagof </span>allow to express dependency of an attribute from an aggregate function of
|
||
another attribute, as in CLP(BN) and PRM, while the possibility of specifying a
|
||
uniform distribution allows the use of the reference uncertainty feature of
|
||
PRM.
|
||
<!--l. 37--><p class="indent" > These extensions are work in progress: they have been implemented but there is
|
||
no paper yet that describes the semantics of the extended language.
|
||
<!--l. 41--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">2 </span> <a
|
||
id="x1-20002"></a>Installation</h3>
|
||
<!--l. 42--><p class="noindent" ><span
|
||
class="cmtt-10">cplint </span>is distributed in source code in the CVS version of Yap. Download it by
|
||
following the instruction in <a
|
||
href="http://www.ncc.up.pt/%7Evsc/Yap/downloads.html" > http://www.ncc.up.pt/%7Evsc/Yap/downloads.html
|
||
</a>.
|
||
|
||
<!--l. 44--><p class="indent" > <span
|
||
class="cmtt-10">cplint </span>requires glu (a subpackage of vis) and glib. You can download glu from <a
|
||
href="http://vlsi.colorado.edu/%7Evis/getting_VIS_2.1.html" >
|
||
http://vlsi.colorado.edu/%7Evis/getting_VIS_2.1.html </a> You can download glib from
|
||
<a
|
||
href="http://www.gtk.org/" > http://www.gtk.org/ </a>. This is a standard GNU package so it is easy to
|
||
install it using the package management software of your Linux or Cygwin
|
||
distribution.
|
||
<!--l. 50--><p class="indent" > Install glu:
|
||
<ol class="enumerate1" >
|
||
<li
|
||
class="enumerate" id="x1-2002x1">downlad <span
|
||
class="cmtt-10">glu-2.1.tar.gz</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2004x2">decompress it
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2006x3"><span
|
||
class="cmtt-10">cd glu-2.1</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2008x4"><span
|
||
class="cmtt-10">mkdir arch</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2010x5"><span
|
||
class="cmtt-10">cd arch</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2012x6"><span
|
||
class="cmtt-10">../configure</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2014x7"><span
|
||
class="cmtt-10">make</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2016x8"><span
|
||
class="cmtt-10">su</span>
|
||
</li>
|
||
<li
|
||
class="enumerate" id="x1-2018x9"><span
|
||
class="cmtt-10">make install</span></li></ol>
|
||
<!--l. 62--><p class="noindent" >This will install glu into <span
|
||
class="cmtt-10">/usr/local</span>, if you want to install to a different <span
|
||
class="cmtt-10">DIR </span>use
|
||
<span
|
||
class="cmtt-10">../configure --prefix DIR</span>
|
||
<!--l. 65--><p class="indent" > Install Yap together with <span
|
||
class="cmtt-10">cplint</span>: when compiling Yap following the instuction of
|
||
the <span
|
||
class="cmtt-10">INSTALL </span>file in the root of the Yap folder, use
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
configure --enable-cplint
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 69--><p class="nopar" > Under Windows, you have to use Cygwin (glu does not compile under MinGW),
|
||
so
|
||
<br class="newline" />
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
configure --enable-cplint --enable-cygwin
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 73--><p class="nopar" > If you installed glu in <span
|
||
class="cmtt-10">DIR</span>, use <span
|
||
class="cmtt-10">--enable-cplint=DIR</span>
|
||
<!--l. 76--><p class="indent" > After having performed <span
|
||
class="cmtt-10">make install </span>you can do <span
|
||
class="cmtt-10">make installcheck </span>that will
|
||
execute a test of <span
|
||
class="cmtt-10">cplint</span>. If no error is reported you have a working installation of
|
||
<span
|
||
class="cmtt-10">cplint</span>.
|
||
<!--l. 79--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">3 </span> <a
|
||
id="x1-30003"></a>Syntax</h3>
|
||
<!--l. 81--><p class="noindent" >Disjunction in the head is represented with a semicolon and atoms in the head are
|
||
separated from probabilities by a colon. For the rest, the usual syntax of Prolog is
|
||
used. For example, the CP-logic clause
|
||
<center class="math-display" >
|
||
<img
|
||
src="manual0x.png" alt="h1 : p1 ∨...∨ hn : pn ← b1,...,bm,<2C>c1,...,<2C>cl " class="math-display" ></center> is
|
||
represented by
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
h1:p1 ; ... ; hn:pn :- b1,...,bm,\+ c1,....,\+ cl
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 87--><p class="nopar" > No parentheses are necessary. The <span
|
||
class="cmtt-10">pi </span>are numeric expressions that can involve
|
||
variables appearing in the body. It is up to the user to ensure that the numeric
|
||
expressions are legal, i.e. that they sum up to less than one for every instantiation of
|
||
the clause for which the body is true in an instance.
|
||
<!--l. 91--><p class="indent" > If the clause has an empty body, it can be represented like this
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
h1:p1 ; ... ;hn:pn.
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 94--><p class="nopar" > If the clause has a single head with probability 1, the annotation can be omitted and
|
||
the clause takes the form of a normal prolog clause, i.e.
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
h1:- b1,...,bm,\+ c1,...,cl.
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 98--><p class="nopar" > stands for
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
h1:1 :- b1,...,bm,\+ c1,...,cl.
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 102--><p class="nopar" >
|
||
<!--l. 104--><p class="indent" > The coin example of <span class="cite">[<a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span> is represented as (see file <span
|
||
class="cmtt-10">coin.cpl</span>)
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
heads(Coin):1/2 ; tails(Coin):1/2:-
|
||
 <br />     toss(Coin),\+biased(Coin).
|
||
 <br />
|
||
 <br />heads(Coin):0.6 ; tails(Coin):0.4:-
|
||
 <br />     toss(Coin),biased(Coin).
|
||
 <br />
|
||
 <br />fair(Coin):0.9 ; biased(Coin):0.1.
|
||
 <br />
|
||
 <br />toss(coin).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 115--><p class="nopar" > The first clause states that if we toss a coin that is not biased it has equal
|
||
probability of landing heads and tails. The second states that if the coin is biased it
|
||
is slightly more probable that it lands heads. The third states that the coin is fair
|
||
with probability 0.9 and biased with probability 0.1 and the last clause states that we
|
||
toss a coin with certainty.
|
||
<!--l. 120--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">4 </span> <a
|
||
id="x1-40004"></a>Commands</h3>
|
||
<!--l. 121--><p class="noindent" >The program must be stored in a text file with extension <span
|
||
class="cmtt-10">.cpl</span>. Suppose
|
||
you have stored the example above in file <span
|
||
class="cmtt-10">coin.cpl</span>. In order to answer
|
||
queries from this program, you have to run <span
|
||
class="cmtt-10">yap</span>, load <span
|
||
class="cmtt-10">cplint </span>by issuing the
|
||
command
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
use_module(library(cplint)).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 126--><p class="nopar" > at the command prompt. Then you must parse the source file <span
|
||
class="cmtt-10">coin.cpl </span>with the
|
||
command
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
p(coin).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 131--><p class="nopar" > if <span
|
||
class="cmtt-10">coin.cpl </span>is in the current directory, or
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
p(’path_to_coin/coin’).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 135--><p class="nopar" > if <span
|
||
class="cmtt-10">coin.cpl </span>is in a different directory. At this point you can pose query to the
|
||
program. You have to use the predicate <span
|
||
class="cmtt-10">s/2 </span>(for solve) that takes as its first
|
||
argument a conjunction of goals in the form of a list and returns the computed
|
||
probability as its second argument. For example, the probability of the conjunction
|
||
<span
|
||
class="cmtt-10">head(coin), biased(coin) </span>can be asked with the query
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
s([head(coin),biased(coin)],P).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 140--><p class="nopar" > For computing the probability of a conjunction given another conjunction you have
|
||
to use the predicate <span
|
||
class="cmtt-10">sc/3 </span>(for solve conditional) that take takes as input the query
|
||
conjunction as its first argument, the evidence conjunction as its second argument
|
||
and returns the probability in its third argument. For example, the probability of the
|
||
query <span
|
||
class="cmtt-10">heads(coin) </span>given the evidence <span
|
||
class="cmtt-10">biased(coin) </span>can be asked with the
|
||
query
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
sc([heads(coin)],[biased(coin)],P).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 145--><p class="nopar" > The package contains also a program <span
|
||
class="cmtt-10">semantics.pl </span>that computes the probability
|
||
of queries by using directly the semantics of LPADs (i.e. by generating all the ground
|
||
instances and then testing the query with each of them). This is provided for testing
|
||
purposes only. After having compiled the program, you can use the same commands
|
||
of <span
|
||
class="cmtt-10">cplint.pl</span>. <span
|
||
class="cmtt-10">semantics.pl </span>requires an extra file in the directory where <span
|
||
class="cmtt-10">coin.cpl</span>
|
||
is: a file with extension <span
|
||
class="cmtt-10">.uni </span>(for universe) that contains, for each variable, the list of
|
||
constants to which the variable can be instantiated. For example, in our case the
|
||
current directory will contain a file <span
|
||
class="cmtt-10">coin.uni </span>that is a Prolog file containing facts of
|
||
the form
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
universe(VarList,ConstList).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 149--><p class="nopar" > where <span
|
||
class="cmtt-10">VarList </span>is a list of variables names (each must be included in single quotes)
|
||
and <span
|
||
class="cmtt-10">ConstList </span>is a list of constants. <span
|
||
class="cmtt-10">semantics.pl </span>generates the grounding by
|
||
instantiating in all possible ways the variables of <span
|
||
class="cmtt-10">VarList </span>with the constants of
|
||
<span
|
||
class="cmtt-10">ConstList</span>. Note that the variables are identified by name, so a variable with
|
||
the same name in two different clauses will be instantiated with the same
|
||
constants.
|
||
<!--l. 152--><p class="indent" > You can test your installation of <span
|
||
class="cmtt-10">cplint </span>by using the program <span
|
||
class="cmtt-10">test.pl</span>: compile
|
||
<span
|
||
class="cmtt-10">cplint.pl</span>, compile <span
|
||
class="cmtt-10">test.pl </span>and exectue the query <span
|
||
class="cmtt-10">t</span>. A number of queries are
|
||
executed against the example programs and the returned probabilities are checked: if
|
||
<span
|
||
class="cmtt-10">t. </span>succeeds, then <span
|
||
class="cmtt-10">cplint </span>is working.
|
||
<!--l. 157--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">5 </span> <a
|
||
id="x1-50005"></a>Extensions</h3>
|
||
<!--l. 158--><p class="noindent" >In this section we will present the extensions to the syntax of LPADs and CP-logic
|
||
programs that <span
|
||
class="cmtt-10">cplint </span>can handle.
|
||
<!--l. 160--><p class="indent" > The first is the use of some standard Prolog predicates. The bodies can contain
|
||
the built-in predicates:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
is/2
|
||
 <br />>/2
|
||
 <br /></2
|
||
 <br />>=/2
|
||
 <br />=</2
|
||
 <br />=:=/2
|
||
 <br />=\=/2
|
||
 <br />true/0
|
||
 <br />false/0
|
||
 <br />=/2
|
||
 <br />==/2
|
||
 <br />\=/2
|
||
 <br />\==/2
|
||
 <br />length/2
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 177--><p class="nopar" > The bodies can also contain the following library predicates:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
member/2
|
||
 <br />max_list/2
|
||
 <br />min_list/2
|
||
 <br />nth0/3
|
||
 <br />nth/3
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 186--><p class="nopar" > plus the predicate
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
average/2
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 190--><p class="nopar" > that, given a list of numbers, computes its arithmetic mean.
|
||
<!--l. 193--><p class="indent" > Moreover, the bodies can contain the predicates <span
|
||
class="cmtt-10">setof/3 </span>and <span
|
||
class="cmtt-10">bagof/3 </span>with the
|
||
same meaning as in Prolog. Existential quantifiers are allowed in both, so for example
|
||
the query
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
setof(Z, (term(X,Y))^foo(X,Y,Z), L).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 196--><p class="nopar" > returns all the instantiations of <span
|
||
class="cmtt-10">Z </span>such that there exists an instantiation of <span
|
||
class="cmtt-10">X </span>and <span
|
||
class="cmtt-10">Y</span>
|
||
for which <span
|
||
class="cmtt-10">foo(X,Y,Z) </span>is true.
|
||
<!--l. 199--><p class="indent" > An example of the use of <span
|
||
class="cmtt-10">setof </span>and <span
|
||
class="cmtt-10">bagof </span>is in the file <span
|
||
class="cmtt-10">female.cpl</span>:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
male(C):M/P ; female(C):F/P:-
|
||
 <br />    person(C),
|
||
 <br />    setof(Male,known_male(Male),LM),
|
||
 <br />    length(LM,M),
|
||
 <br />    setof(Female,known_female(Female),LF),
|
||
 <br />    length(LF,F),
|
||
 <br />    P is F+M.
|
||
 <br />
|
||
 <br />person(f).
|
||
 <br />
|
||
 <br />known_female(a).
|
||
 <br />
|
||
 <br />known_female(b).
|
||
 <br />
|
||
 <br />known_female(c).
|
||
 <br />
|
||
 <br />known_male(d).
|
||
 <br />
|
||
 <br />known_male(e).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 220--><p class="nopar" > The disjunctive rule expresses the probability of a person of unknown sex of being
|
||
male or female depending on the number of males and females that are known. This
|
||
is an example of the use of expressions in the probabilities in the head that depend
|
||
on variables in the body. The probabilities are well defined because they always sum
|
||
to 1 (unless <span
|
||
class="cmtt-10">P </span>is 0).
|
||
<!--l. 224--><p class="indent" > Another use of <span
|
||
class="cmtt-10">setof </span>and <span
|
||
class="cmtt-10">bagof </span>is to have an attribute depend on an
|
||
aggregate function of another attribute, similarly to what is done in PRM and
|
||
CLP(BN).
|
||
<!--l. 226--><p class="indent" > So, in the classical school example (available in <span
|
||
class="cmtt-10">student.cpl</span>) you can find the
|
||
following clauses:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
student_rank(S,h):0.6 ; student_rank(S,l):0.4:-
|
||
 <br />    bagof(G,R^(registr_stu(R,S),registr_gr(R,G)),L),
|
||
 <br />    average(L,Av),Av>1.5.
|
||
 <br />
|
||
 <br />student_rank(S,h):0.4 ; student_rank(S,l):0.6:-
|
||
 <br />    bagof(G,R^(registr_stu(R,S),registr_gr(R,G)),L),
|
||
 <br />    average(L,Av),Av =< 1.5.
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 236--><p class="nopar" > where <span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">registr_stu(R,S)</span></span></span> expresses that registration <span
|
||
class="cmtt-10">R </span>refers to student <span
|
||
class="cmtt-10">S </span>and
|
||
<span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">registr_gr(R,G)</span></span></span> expresses that registration <span
|
||
class="cmtt-10">R </span>reports grade <span
|
||
class="cmtt-10">G </span>which is a natural
|
||
number. The two clauses express a dependency of the rank of the student from the
|
||
average of her grades.
|
||
<!--l. 239--><p class="indent" > Another extension has been introduced in order to be able to represent reference
|
||
uncertainty of PRMs. Reference uncertainty means that the link structure of a
|
||
relational model is not fixed but is uncertain: this is represented by having the
|
||
instance referenced in a relationship be chosen uniformly from a set. For example,
|
||
consider a domain modeling scientific papers: you have a single entity, paper, and a
|
||
relationship, cites, between paper and itself that connects the citing paper to the
|
||
cited paper. To represent the fact that the cited paper and the citing paper are
|
||
selected uniformly from certain sets, the following clauses can be used (see file
|
||
<span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">paper_ref_simple.cpl</span></span></span>):
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
uniform(cites_cited(C,P),P,L):-
|
||
 <br />    bagof(Pap,paper_topic(Pap,theory),L).
|
||
 <br />
|
||
 <br />uniform(cites_citing(C,P),P,L):-
|
||
 <br />    bagof(Pap,paper_topic(Pap,ai),L).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 246--><p class="nopar" > The first clauses states that the paper <span
|
||
class="cmtt-10">P </span>cited in a citation <span
|
||
class="cmtt-10">C </span>is selected
|
||
uniformly from the set of all papers with topic theory. The second clauses
|
||
expresses that the citing paper is selected uniformly from the papers with topic
|
||
ai.
|
||
<!--l. 251--><p class="indent" > These clauses make use of the predicate
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
uniform(Atom,Variable,List)
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 254--><p class="nopar" > in the head, where <span
|
||
class="cmtt-10">Atom </span>must contain <span
|
||
class="cmtt-10">Variable</span>. The meaning is the following:
|
||
the set of all the atoms obtained by instantiating <span
|
||
class="cmtt-10">Variable </span>of <span
|
||
class="cmtt-10">Atom </span>with a
|
||
term taken from <span
|
||
class="cmtt-10">List </span>is generated and the head is obtained by having a
|
||
disjunct for each instantiation with probability 1<span
|
||
class="cmmi-10">∕N </span>where <span
|
||
class="cmmi-10">N </span>is the length of
|
||
<span
|
||
class="cmtt-10">List</span>.
|
||
<!--l. 258--><p class="indent" > A more elaborate example is present in file <span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">paper_ref.cpl</span></span></span>:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
uniform(cites_citing(C,P),P,L):-
|
||
 <br />    setof(Pap,paper(Pap),L).
|
||
 <br />
|
||
 <br />cites_cited_group(C,theory):0.9 ; cites_cited_group(C,ai):0.1:-
|
||
 <br />    cites_citing(C,P),paper_topic(P,theory).
|
||
 <br />
|
||
 <br />cites_cited_group(C,theory):0.01;cites_cited_group(C,ai):0.99:-
|
||
 <br />    cites_citing(C,P),paper_topic(P,ai).
|
||
 <br />
|
||
 <br />uniform(cites_cited(C,P),P,L):-
|
||
 <br />    cites_cited_group(C,T),bagof(Pap,paper_topic(Pap,T),L).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 271--><p class="nopar" > where the cited paper depends on the topic of the citing paper. In particular, if the
|
||
topic is theory, the cited paper is selected uniformly from the papers about theory
|
||
with probability 0.9 and from the papers about ai with probability 0.1. if
|
||
the topic is ai, the cited paper is selected uniformly from the papers about
|
||
theory with probability 0.01 and from the papers about ai with probability
|
||
0.99.
|
||
<!--l. 274--><p class="indent" > PRMs take into account as well existence uncertainty, where the existence of
|
||
instances is also probabilistic. For example, in the paper domain, the total number of
|
||
citations may be unknown and a citation between any two paper may have a
|
||
probability of existing. For example, a citation between two paper may be more
|
||
probable if they are about the same topic:
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
cites(X,Y):0.005 :-
|
||
 <br />    paper_topic(X,theory),paper_topic(Y,theory).
|
||
 <br />
|
||
 <br />cites(X,Y):0.001 :-
|
||
 <br />    paper_topic(X,theory),paper_topic(Y,ai).
|
||
 <br />
|
||
 <br />cites(X,Y):0.003 :-
|
||
 <br />    paper_topic(X,ai),paper_topic(Y,theory).
|
||
 <br />
|
||
 <br />cites(X,Y):0.008 :-
|
||
 <br />    paper_topic(X,ai),paper_topic(Y,ai).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 287--><p class="nopar" > This is an example of a CP-logic program, because the probabilities in the head do
|
||
not sum up to one. The first clause states that, if the topic of a paper <span
|
||
class="cmtt-10">X </span>is
|
||
theory and of paper <span
|
||
class="cmtt-10">Y </span>is theory, there is a probability of 0.005 that there is a
|
||
citation from <span
|
||
class="cmtt-10">X </span>to <span
|
||
class="cmtt-10">Y</span>. The other clauses consider the remaining cases for the
|
||
topics.
|
||
<!--l. 291--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">6 </span> <a
|
||
id="x1-60006"></a>Parameters</h3>
|
||
<!--l. 292--><p class="noindent" >The proof procedure has two parameters that can be set with the command
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
set(parameter,value).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 295--><p class="nopar" > from the Yap prompt after having compiled <span
|
||
class="cmtt-10">cplint.pl</span>. The current value can be
|
||
read with
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
setting(parameter,Value).
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 300--><p class="nopar" > from the Yap prompt. Available parameters:
|
||
<ul class="itemize1">
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">epsilon_parsing</span></span></span>: if (1 - the sum of the probabilities of all the head
|
||
atoms) is smaller than <span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">epsilon_parsing</span></span></span> then the clause is considered as
|
||
an LPAD clause, otherwise it is considered as a CP-logic clause. Default
|
||
value 0.00001
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">savedot</span>: if true a graph representing the BDD is saved in the file <span
|
||
class="cmtt-10">cpl.dot </span>in
|
||
the current directory in dot format. The variables names are of the
|
||
form <span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">Xn_m</span></span></span> where <span
|
||
class="cmtt-10">n </span>is the number of the multivalued variable and <span
|
||
class="cmtt-10">m </span>is
|
||
the number of the binary variable. The correspondence of variables
|
||
to clauses can be evinced from the list printed on the screen of the
|
||
form
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
Variables: [(2,[X=2,X1=1]),(2,[X=1,X1=0]),(1,[])]
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 315--><p class="nopar" > where the first element of each couple is the clause number of the input file
|
||
(starting from 1). In the example above variable <span
|
||
class="cmtt-10">X0 </span>corresponds to clause <span
|
||
class="cmtt-10">2</span>
|
||
with the substitutions <span
|
||
class="cmtt-10">X=2,X1=1</span>, variable <span
|
||
class="cmtt-10">X1 </span>corresponds to clause <span
|
||
class="cmtt-10">2 </span>with the
|
||
substitutions <span
|
||
class="cmtt-10">X=1,X1=0 </span>and variable <span
|
||
class="cmtt-10">X2 </span>corresponds to clause <span
|
||
class="cmtt-10">1 </span>with the empty
|
||
substitution. You can view the graph with <span
|
||
class="cmtt-10">graphviz </span>(<a
|
||
href="www.graphviz.org" > www.graphviz.org </a>)
|
||
using the command
|
||
|
||
<table
|
||
class="verbatim"><tr class="verbatim"><td
|
||
class="verbatim"><div class="verbatim">
|
||
dotty cpl.dot &
|
||
</div>
|
||
</td></tr></table>
|
||
<!--l. 324--><p class="nopar" ></li></ul>
|
||
<!--l. 328--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">7 </span> <a
|
||
id="x1-70007"></a>Additional Files</h3>
|
||
<!--l. 329--><p class="noindent" >In the directory where Yap keeps the library files (usually <span
|
||
class="cmtt-10">/usr/local/share/Yap</span>)
|
||
you can find the directory <span
|
||
class="cmtt-10">cplint </span>that contains additional files. <span
|
||
class="cmtt-10">cplint</span>
|
||
contains
|
||
<ul class="itemize1">
|
||
<li class="itemize"><span
|
||
class="cmtt-10">semantics.pl</span>: Prolog program for computing the probability according
|
||
to the semantics.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">test.pl</span>: Prolog program for testing the system.
|
||
</li>
|
||
<li class="itemize">Subdirectory <span
|
||
class="cmtt-10">examples</span>:
|
||
<ul class="itemize2">
|
||
<li class="itemize"><span
|
||
class="cmtt-10">alarm.cpl</span>: representation of the Bayesian network in Figure 2 of
|
||
<span class="cite">[<a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">coin.cpl</span>: coin example from <span class="cite">[<a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">coin2.cpl</span>: coin example with two coins.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">dice.cpl</span>: dice example from <span class="cite">[<a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">twosideddice.cpl,threesideddice.cpl</span></span></span> game with idealized dice
|
||
with two or three sides. Used in the experiments in <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">es.cpl</span>: first example in <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">esapprox.cpl</span>: example showing the problems of approximate
|
||
inference (see <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span>).
|
||
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">esrange.cpl</span>: example showing the problems with non range
|
||
restricted programs (see <span class="cite">[<a
|
||
href="#XRig-RCRA07-IC">7</a>]</span>).
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">female.cpl</span>: example showing the dependence of probabilities in the
|
||
head from variables in the body (from <span class="cite">[<a
|
||
href="#XVenVer04-ICLP04-IC">11</a>]</span>).
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">mendel.cpl</span>: program describing the Mendelian rules of inheritance,
|
||
taken from <span class="cite">[<a
|
||
href="#XBlo04-ILP04WIP-IC">2</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">paper_ref.cpl,paper_ref_simple.cpl</span></span></span>: paper citations examples,
|
||
showing reference uncertainty, inspired by <span class="cite">[<a
|
||
href="#XGetoor+al:JMLR02">5</a>]</span>.
|
||
</li>
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">paper_ref_not.cpl</span></span></span>: paper citations example showing that negation
|
||
can be used also for predicates defined by clauses with <span
|
||
class="cmtt-10">uniform </span>in
|
||
the head.
|
||
</li>
|
||
<li class="itemize"><span
|
||
class="cmtt-10">school.cpl</span>: example inspired by the example <span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">school_32.yap</span></span></span> from
|
||
the source distribution of Yap in the <span
|
||
class="cmtt-10">CLPBN </span>directory.
|
||
</li>
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">school_simple.cpl</span></span></span>: simplified version of <span
|
||
class="cmtt-10">school.cpl</span>.
|
||
</li>
|
||
<li class="itemize"><span class="obeylines-h"><span class="verb"><span
|
||
class="cmtt-10">student.cpl</span></span></span>: student example from Figure 1.3 of <span class="cite">[<a
|
||
href="#XGetFri01-BC">4</a>]</span>.</li></ul>
|
||
<!--l. 354--><p class="noindent" >The files <span
|
||
class="cmtt-10">*.uni </span>that are present for some of the examples are used by
|
||
<span
|
||
class="cmtt-10">semantics.pl</span>. Some of the example files contain in an initial comment some
|
||
queries together with their result.
|
||
</li>
|
||
<li class="itemize">Subdirectory <span
|
||
class="cmtt-10">doc</span>: contains this manual in latex, html and pdf.</li></ul>
|
||
<!--l. 357--><p class="noindent" >
|
||
<h3 class="sectionHead"><span class="titlemark">8 </span> <a
|
||
id="x1-80008"></a>License</h3>
|
||
<!--l. 362--><p class="noindent" ><span
|
||
class="cmtt-10">cplint</span>, as Yap, follows the Artistic License 2.0 that you can find in Yap CVS root
|
||
dir. The copyright is by Fabrizio Riguzzi.
|
||
<!--l. 367--><p class="indent" > The program uses the library <a
|
||
href="http://vlsi.colorado.edu/%7Efabio/" > CUDD </a> for manipulating BDDs that is included in
|
||
glu. For the use of CUDD, the following license must be accepted:
|
||
<!--l. 372--><p class="indent" > Copyright (c) 1995-2004, Regents of the University of Colorado
|
||
<!--l. 374--><p class="indent" > All rights reserved.
|
||
<!--l. 376--><p class="indent" > Redistribution and use in source and binary forms, with or without modification,
|
||
are permitted provided that the following conditions are met:
|
||
|
||
<ul class="itemize1">
|
||
<li class="itemize">Redistributions of source code must retain the above copyright notice, this
|
||
list of conditions and the following disclaimer.
|
||
</li>
|
||
<li class="itemize">Redistributions in binary form must reproduce the above copyright notice,
|
||
this list of conditions and the following disclaimer in the documentation
|
||
and/or other materials provided with the distribution.
|
||
</li>
|
||
<li class="itemize">Neither the name of the University of Colorado nor the names of its
|
||
contributors may be used to endorse or promote products derived from
|
||
this software without specific prior written permission.</li></ul>
|
||
<!--l. 393--><p class="noindent" >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 PURPOSE ARE
|
||
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
|
||
CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
||
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
|
||
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
||
OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
|
||
INTERRUPTION) HOWEVER CAUSED 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.
|
||
<!--l. 1--><p class="noindent" >
|
||
<h3 class="likesectionHead"><a
|
||
id="x1-90008"></a>References</h3>
|
||
<!--l. 1--><p class="noindent" >
|
||
<div class="thebibliography">
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[1]<span class="bibsp">   </span></span><a
|
||
id="XDBLP:journals/ngc/AptB91"></a>K. R. Apt and M. Bezem. Acyclic programs. <span
|
||
class="cmti-10">New Generation</span>
|
||
<span
|
||
class="cmti-10">Comput.</span>, 9(3/4):335–364, 1991.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[2]<span class="bibsp">   </span></span><a
|
||
id="XBlo04-ILP04WIP-IC"></a>H. Blockeel. Probabilistic logical models for mendel’s experiments: An
|
||
exercise. In <span
|
||
class="cmti-10">Inductive Logic Programming (ILP 2004), Work in Progress</span>
|
||
<span
|
||
class="cmti-10">Track</span>, 2004.
|
||
</p>
|
||
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[3]<span class="bibsp">   </span></span><a
|
||
id="XDBLP:conf/ijcai/RaedtKT07"></a>L. De Raedt, A. Kimmig, and H. Toivonen. Problog: A probabilistic
|
||
prolog and its application in link discovery. In <span
|
||
class="cmti-10">Proceedings of the 20th</span>
|
||
<span
|
||
class="cmti-10">International Joint Conference on Artificial Intelligence</span>, pages 2462–2467,
|
||
2007.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[4]<span class="bibsp">   </span></span><a
|
||
id="XGetFri01-BC"></a>L. Getoor, N. Friedman, D. Koller, and A. Pfeffer. Learning
|
||
probabilistic relational models. In Saso Dzeroski and Nada Lavrac, editors,
|
||
<span
|
||
class="cmti-10">Relational Data Mining</span>. Springer-Verlag, Berlin, 2001.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[5]<span class="bibsp">   </span></span><a
|
||
id="XGetoor+al:JMLR02"></a>L. Getoor, N. Friedman, D. Koller, and B. Taskar. Learning
|
||
probabilistic models of relational structure. <span
|
||
class="cmti-10">Journal of Machine Learning</span>
|
||
<span
|
||
class="cmti-10">Research</span>, 3:679–707, December 2002.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[6]<span class="bibsp">   </span></span><a
|
||
id="XRig-AIIA07-IC"></a>Fabrizio Riguzzi. A top down interpreter for lpad and cp-logic. In
|
||
<span
|
||
class="cmti-10">10th Congress of the Italian Association for Artificial Intelligence</span>. Springer,
|
||
2007.
|
||
<a
|
||
href="http://www.ing.unife.it/docenti/FabrizioRiguzzi/Papers/Rig-AIIA07.pdf" > http://www.ing.unife.it/docenti/FabrizioRiguzzi/Papers/Rig-AIIA07.pdf
|
||
</a>.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[7]<span class="bibsp">   </span></span><a
|
||
id="XRig-RCRA07-IC"></a>Fabrizio Riguzzi. A top down interpreter for lpad and
|
||
cp-logic. In <span
|
||
class="cmti-10">The 14th RCRA workshop Experimental Evaluation of</span>
|
||
<span
|
||
class="cmti-10">Algorithms for Solving Problems with Combinatorial Explosion</span>, 2007. <a
|
||
href="http://pst.istc.cnr.it/RCRA07/articoli/P19-riguzzi-RCRA07.pdf" >
|
||
http://pst.istc.cnr.it/RCRA07/articoli/P19-riguzzi-RCRA07.pdf </a>.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[8]<span class="bibsp">   </span></span><a
|
||
id="XSanPagQaz03-UAI-IC"></a>V. Santos Costa, D. Page, M. Qazi, and J. Cussens. CLP(BN):
|
||
Constraint logic programming for probabilistic knowledge. In <span
|
||
class="cmti-10">Uncertainty</span>
|
||
<span
|
||
class="cmti-10">in Artificial Intelligence (UAI 2003)</span>, 2003.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[9]<span class="bibsp">   </span></span><a
|
||
id="XVenDenBru-JELIA06"></a>J. Vennekens, M. Denecker, and M. Bruynooghe. Representing causal
|
||
information about a probabilistic process. In <span
|
||
class="cmti-10">10th European Conference on</span>
|
||
<span
|
||
class="cmti-10">Logics in Artificial Intelligence, JELIA 2006</span>, LNAI. Springer, September
|
||
2006.
|
||
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[10]<span class="bibsp">   </span></span><a
|
||
id="XVenVer03-TR"></a>J. Vennekens and S. Verbaeten. Logic programs with annotated
|
||
disjunctions. Technical Report CW386, K. U. Leuven, 2003. <a
|
||
href="http://www.cs.kuleuven.ac.be/%7Ejoost/techrep.ps" >
|
||
http://www.cs.kuleuven.ac.be/%7Ejoost/techrep.ps </a>.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[11]<span class="bibsp">   </span></span><a
|
||
id="XVenVer04-ICLP04-IC"></a>J. Vennekens, S. Verbaeten,
|
||
and M. Bruynooghe. Logic programs with annotated disjunctions. In <span
|
||
class="cmti-10">The</span>
|
||
<span
|
||
class="cmti-10">20th International Conference on Logic Programming (ICLP 2004)</span>, 2004. <a
|
||
href="http://www.cs.kuleuven.ac.be/%7Ejoost/" >
|
||
http://www.cs.kuleuven.ac.be/%7Ejoost/ </a>.
|
||
</p>
|
||
<p class="bibitem" ><span class="biblabel">
|
||
[12]<span class="bibsp">   </span></span><a
|
||
id="XCP-logic-unp"></a>Joost Vennekens, Marc Denecker, and Maurice Bruynooge. Extending
|
||
the role of causality in probabilistic modeling. <a
|
||
href="http://www.cs.kuleuven.ac.be/%7Ejoost/cplogic.pdf" >
|
||
http://www.cs.kuleuven.ac.be/%7Ejoost/cplogic.pdf </a>, 2006.
|
||
</p>
|
||
</div>
|
||
|
||
</body></html>
|
||
|
||
|
||
|