Snippet of code

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
/*
 *  Copyright or © or Copr. Ecole des Mines d'Alès (2012-2014)
 * 
 *  This software is a computer program whose purpose is to provide
 *  several functionalities for the processing of semantic data
 *  sources such as ontologies or text corpora.
 * 
 *  This software is governed by the CeCILL  license under French law and
 *  abiding by the rules of distribution of free software.  You can  use,
 *  modify and/ or redistribute the software under the terms of the CeCILL
 *  license as circulated by CEA, CNRS and INRIA at the following URL
 *  "http://www.cecill.info".
 *
 *  As a counterpart to the access to the source code and  rights to copy,
 *  modify and redistribute granted by the license, users are provided only
 *  with a limited warranty  and the software's author,  the holder of the
 *  economic rights,  and the successive licensors  have only  limited
 *  liability.
 
 *  In this respect, the user's attention is drawn to the risks associated
 *  with loading,  using,  modifying and/or developing or reproducing the
 *  software by the user in light of its specific status of free software,
 *  that may mean  that it is complicated to manipulate,  and  that  also
 *  therefore means  that it is reserved for developers  and  experienced
 *  professionals having in-depth computer knowledge. Users are therefore
 *  encouraged to load and test the software's suitability as regards their
 *  requirements in conditions enabling the security of their systems and/or
 *  data to be ensured and,  more generally, to use and operate it in the
 *  same conditions as regards security.
 *
 *  The fact that you are presently reading this means that you have had
 *  knowledge of the CeCILL license and that you accept its terms.
 */
package slib.examples.sml.yago;
 
import java.util.ArrayList;
import java.util.List;
import java.util.Random;
import java.util.Set;
import org.openrdf.model.URI;
import slib.graph.algo.utils.GAction;
import slib.graph.algo.utils.GActionType;
import slib.graph.algo.validator.dag.ValidatorDAG;
import slib.graph.io.conf.GDataConf;
import slib.graph.io.conf.GraphConf;
import slib.graph.io.loader.GraphLoaderGeneric;
import slib.graph.io.util.GFormat;
import slib.graph.model.graph.G;
import slib.graph.model.impl.graph.memory.GraphMemory;
import slib.graph.model.impl.repo.URIFactoryMemory;
import slib.graph.model.repo.URIFactory;
import slib.sml.sm.core.engine.SM_Engine;
import slib.sml.sm.core.metrics.ic.utils.IC_Conf_Topo;
import slib.sml.sm.core.metrics.ic.utils.ICconf;
import slib.sml.sm.core.utils.SMConstants;
import slib.sml.sm.core.utils.SMconf;
import slib.utils.ex.SLIB_Exception;
import slib.utils.impl.Timer;
 
/**
 *
 * @author Sébastien Harispe <sebastien.harispe@gmail.com>
 */
public class SMComputationYago {
     
    public static void main(String[] args) throws SLIB_Exception{
         
        /*
         * The data loading is composed of three steps:
         *  - (1) we load the yago taxonomy from the turtle file
         *  - (2) we type the vertices in order to specify the engine which are the vertices associated to classes (concepts)
         *       This treatment is required in order to perform some algorithms.
         *  - (3) We root vertices which are not rooted by owl:Thing. Algorithms require the processed graph to be connected
         *  i.e. to compute the Most Informative Common Ancestors of two concepts.
         *
         * Notice that due to the size of the taxonomy, extra memory must be allocated to the JVM e.g. -Xmx3000m
         */
         
        Timer t = new Timer();
        t.start();
         
        String yagoTaxonomyFile = "/data/yago/yagoTaxonomy.ttl";
         
        URIFactory factory = URIFactoryMemory.getSingleton();
        URI yagoURI = factory.getURI("http://yago-knowledge.org/resource/");
        G g = new GraphMemory(yagoURI);
         
        // This is the configuration of the data
        GDataConf dataConf = new GDataConf(GFormat.TURTLE, yagoTaxonomyFile);
         
         
        // We specify an action to root the vertices, typed as class without outgoing rdfs:subclassOf relationship
        // Those vertices are linked to owl:Thing by an eddge x  rdfs:subClassOf owl:Thing
        GAction actionRerootConf = new GAction(GActionType.REROOTING);
         
        // We now create the configuration we will specify to the generic loader
        GraphConf gConf = new GraphConf();
        gConf.addGDataConf(dataConf);
        gConf.addGAction(actionRerootConf);
         
        GraphLoaderGeneric.load(gConf,g);
         
        System.out.println(g.toString());
         
        // The taxonomy is now a rDAG, i.e. rooted Directed Acyclic Graph.
        // Check by yourself
        Set<URI> roots = new ValidatorDAG().getTaxonomicRoots(g);
        System.out.println("Roots: "+roots);
         
         
        // We compute the similarity between two concepts
        URI wikiRugbyFoorballerURI = factory.getURI(yagoURI.stringValue() + "wikicategory_Rugby_footballers");
        URI WordnetSoccerPlayerURI = factory.getURI(yagoURI.stringValue() + "wordnet_soccer_player_110618342");
 
        // First we configure an intrincic IC
        ICconf icConf = new IC_Conf_Topo(SMConstants.FLAG_ICI_DEPTH_MAX_NONLINEAR);
        // Then we configure the pairwise measure to use, we here choose to use Lin formula
        SMconf smConf = new SMconf(SMConstants.FLAG_SIM_PAIRWISE_DAG_NODE_LIN_1998, icConf);
 
        // We define the engine used to compute the similarity
        SM_Engine engine = new SM_Engine(g);
 
 
        double sim = engine.compare(smConf, wikiRugbyFoorballerURI, WordnetSoccerPlayerURI);
        System.out.println("Similarity: " + sim);
 
        /*
         * Notice that the first computation is expensive as the engine compute the IC and extra information
         * which are cached by the engine
         * Let's perform 10000 random computations (we only print some results).
         * We retrieve the set of vertices as a list
         */
        int totalComparison = 10000;
 
        List<URI> listVertices = new ArrayList<URI>(g.getV());
        int nbConcepts = listVertices.size();
        int id1, id2;
        URI c1, c2;
        String idC1, idC2;
        Random r = new Random();
 
        for (int i = 0; i < totalComparison; i++) {
            id1 = r.nextInt(nbConcepts);
            id2 = r.nextInt(nbConcepts);
 
            c1 = listVertices.get(id1);
            c2 = listVertices.get(id2);
 
            sim = engine.compare(smConf, c1, c2);
 
            if ((i + 1) % 1000 == 0) {
                idC1 = c1.getLocalName();
                idC2 = c2.getLocalName();
 
                System.out.println("Sim " + (i + 1) + "/" + totalComparison + "\t" + idC1 + "/" + idC2 + ": " + sim);
            }
        }
        t.stop();
        t.elapsedTime();
         
    }
     
}