Package it.unimi.dsi.mg4j.query

Examples of it.unimi.dsi.mg4j.query.QueryEngine

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
   
    /* To run a query in a simple way we need a query engine. The engine requires a parser
     * (which in turn requires the set of index names and a default index), a document iterator
     * builder, which needs the index map, a default index, and a limit on prefix query
     * expansion, and finally the index map. */
    QueryEngine engine = new QueryEngine(
      new SimpleParser( indexMap.keySet(), "text", termProcessors ),
      new DocumentIteratorBuilderVisitor( indexMap, text, 1000 ),
      indexMap
     
    );

    /* Optionally, we can score the results. Here we use a state-of-art ranking
     * function, BM25, which requires document sizes. */
    engine.score( new BM25Scorer() );
   
    /* Optionally, we can weight the importance of each index. To do so, we have to pass a map,
     * and again we use the handy fastutil constructor. Note that setting up a BM25F scorer
     * would give much better results, but we want to keep it simple. */
    engine.setWeights( new Reference2DoubleOpenHashMap<Index>( new Index[] { text, title }, new double[] { 1, 2 } ) );
   
    /* Optionally, we can use an interval selector to get intervals representing matches. */
    engine.intervalSelector = new IntervalSelector();
   
    /* We are ready to run our query. We just need a list to store its results. The list is made
     * of DocumentScoreInfo objects, which comprise a document id, a score, and possibly an
     * info field that is generic. Here the info field is a map from indices to arrays
     * of selected intervals. This part will be empty if we do not set an interval selector. */
    ObjectArrayList<DocumentScoreInfo<Reference2ObjectMap<Index, SelectedInterval[]>>> result =
      new ObjectArrayList<DocumentScoreInfo<Reference2ObjectMap<Index,SelectedInterval[]>>>();

    /* The query engine can return any subsegment of the results of a query. Here we grab the first 20 results. */
    engine.process( arg[ 1 ], 0, 20, result );
   
    for( DocumentScoreInfo<Reference2ObjectMap<Index, SelectedInterval[]>> dsi : result ) {
      System.out.println( dsi.document + " " + dsi.score );
    }
  }
View Full Code Here
TOP

Related Classes of it.unimi.dsi.mg4j.query.QueryEngine

  • it.unimi.dsi.fastutil.ints.IntOpenHashSet
  • it.unimi.dsi.mg4j.examples.RunQuery
  • it.unimi.dsi.mg4j.search.DocumentIterator
  • it.unimi.dsi.mg4j.search.score.LinearAggregator
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.