Package org.apache.lucene.search.similarities

Examples of org.apache.lucene.search.similarities.Similarity
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
  public void testSweetSpotComputeNorm() {
 
    final SweetSpotSimilarity ss = new SweetSpotSimilarity();
    ss.setLengthNormFactors(1,1,0.5f,true);

    Similarity d = new DefaultSimilarity();
    Similarity s = ss;


    // base case, should degrade
    FieldInvertState invertState = new FieldInvertState("bogus");
    invertState.setBoost(1.0f);
    for (int i = 1; i < 1000; i++) {
      invertState.setLength(i);
      assertEquals("base case: i="+i,
                   computeAndGetNorm(d, invertState),
                   computeAndGetNorm(s, invertState),
                   0.0f);
    }

    // make a sweet spot
 
    ss.setLengthNormFactors(3,10,0.5f,true);
 
    for (int i = 3; i <=10; i++) {
      invertState.setLength(i);
      assertEquals("3,10: spot i="+i,
                   1.0f,
                   computeAndDecodeNorm(ss, ss, invertState),
                   0.0f);
    }
 
    for (int i = 10; i < 1000; i++) {
      invertState.setLength(i-9);
      final byte normD = computeAndGetNorm(d, invertState);
      invertState.setLength(i);
      final byte normS = computeAndGetNorm(s, invertState);
      assertEquals("3,10: 10<x : i="+i,
                   normD,
                   normS,
                   0.0f);
    }


    // separate sweet spot for certain fields

    final SweetSpotSimilarity ssBar = new SweetSpotSimilarity();
    ssBar.setLengthNormFactors(8,13, 0.5f, false);
    final SweetSpotSimilarity ssYak = new SweetSpotSimilarity();
    ssYak.setLengthNormFactors(6,9, 0.5f, false);
    final SweetSpotSimilarity ssA = new SweetSpotSimilarity();
    ssA.setLengthNormFactors(5,8,0.5f, false);
    final SweetSpotSimilarity ssB = new SweetSpotSimilarity();
    ssB.setLengthNormFactors(5,8,0.1f, false);
   
    Similarity sp = new PerFieldSimilarityWrapper() {
      @Override
      public Similarity get(String field) {
        if (field.equals("bar"))
          return ssBar;
        else if (field.equals("yak"))
View Full Code Here
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
    IndexReader ir = iw.getReader();
    iw.close();
   
    // no boosting
    IndexSearcher searcher1 = newSearcher(ir, false);
    final Similarity base = searcher1.getSimilarity();
    // boosting
    IndexSearcher searcher2 = newSearcher(ir, false);
    searcher2.setSimilarity(new PerFieldSimilarityWrapper() {
      final Similarity fooSim = new BoostingSimilarity(base, "foo_boost");

      @Override
      public Similarity get(String field) {
        return "foo".equals(field) ? fooSim : base;
      }

      @Override
      public float coord(int overlap, int maxOverlap) {
        return base.coord(overlap, maxOverlap);
      }

      @Override
      public float queryNorm(float sumOfSquaredWeights) {
        return base.queryNorm(sumOfSquaredWeights);
      }
    });
   
    // in this case, we searched on field "foo". first document should have 2x the score.
    TermQuery tq = new TermQuery(new Term("foo", "quick"));
View Full Code Here
99
100
101
102
103
104
105
106
107
108
109
  }
 
  @Override
  public synchronized Similarity get(String field) {
    assert field != null;
    Similarity sim = previousMappings.get(field);
    if (sim == null) {
      sim = knownSims.get(Math.max(0, Math.abs(perFieldSeed ^ field.hashCode())) % knownSims.size());
      previousMappings.put(field, sim);
    }
    return sim;
View Full Code Here
79
80
81
82
83
84
85
86
87
88
89
    dir.close();
    dir = null;
  }

  public void testNorm() throws Exception {
    Similarity saved = searcher.getSimilarity();
    try {
      // no norm field (so agnostic to indexed similarity)
      searcher.setSimilarity(sim);
      assertHits(new FunctionQuery(
          new NormValueSource("text")),
View Full Code Here
180
181
182
183
184
185
186
187
188
189
190
    assertHits(new FunctionQuery(new FloatFieldSource("float")),
        new float[] { 5.2f, 9.3f });
  }
 
  public void testIDF() throws Exception {
    Similarity saved = searcher.getSimilarity();
    try {
      searcher.setSimilarity(new DefaultSimilarity());
      assertHits(new FunctionQuery(
          new IDFValueSource("bogus", "bogus", "text", new BytesRef("test"))),
          new float[] { 0.5945349f, 0.5945349f });
View Full Code Here
245
246
247
248
249
250
251
252
253
254
255
        new ConstValueSource(1f), new ConstValueSource(2f)})),
        new float[] { 1f, 1f });
  }
 
  public void testNorm() throws Exception {
    Similarity saved = searcher.getSimilarity();
    try {
      // no norm field (so agnostic to indexed similarity)
      searcher.setSimilarity(new DefaultSimilarity());
      assertHits(new FunctionQuery(
          new NormValueSource("byte")),
View Full Code Here
332
333
334
335
336
337
338
339
340
341
342
        new TermFreqValueSource("bogus", "bogus", "string", new BytesRef("bar"))),
        new float[] { 0f, 1f });
  }
 
  public void testTF() throws Exception {
    Similarity saved = searcher.getSimilarity();
    try {
      // no norm field (so agnostic to indexed similarity)
      searcher.setSimilarity(new DefaultSimilarity());
      assertHits(new FunctionQuery(
          new TFValueSource("bogus", "bogus", "text", new BytesRef("test"))),
View Full Code Here
238
239
240
241
242
243
244
245
246
247
248
    query.add(new TermQuery(new Term(field, "xx")), BooleanClause.Occur.MUST);
    query.add(new TermQuery(new Term(field, "w2")), BooleanClause.Occur.MUST);
    query.add(new TermQuery(new Term(field, "zz")), BooleanClause.Occur.SHOULD);

    int[] expDocNrs = {2, 3};
    Similarity oldSimilarity = searcher.getSimilarity();
    try {
      searcher.setSimilarity(new DefaultSimilarity(){
        @Override
        public float coord(int overlap, int maxOverlap) {
          return overlap / ((float)maxOverlap - 1);
View Full Code Here
55
56
57
58
59
60
61
62
63
64
65
    NamedList<Double> paramNamedList = new NamedList<Double>();
    paramNamedList.add("mu", 2500.0);
    SolrParams params = SolrParams.toSolrParams(paramNamedList);
    LMDirichletSimilarityFactory factory = new LMDirichletSimilarityFactory();
    factory.init(params);
    Similarity simLMDir = factory.getSimilarity();
    searcher.setSimilarity(simLMDir);

    parserArticle = new QueryParser(Version.LUCENE_43, IndexField.TEXT.name, IndexWikipediaDump.ANALYZER);
    parserTitle = new QueryParser(Version.LUCENE_43, IndexField.TITLE.name, IndexWikipediaDump.ANALYZER);
  }
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
    IndexReader reader = DirectoryReader.open(FSDirectory.open(indexLocation));
    IndexSearcher searcher = new IndexSearcher(reader);

    if (similarity.equalsIgnoreCase("BM25")) {
      Similarity simBM25 = new BM25Similarity();
      searcher.setSimilarity(simBM25);
    } else if (similarity.equalsIgnoreCase("LM")) {
      NamedList<Double> paramNamedList = new NamedList<Double>();
      paramNamedList.add("mu", 2500.0);
      SolrParams params = SolrParams.toSolrParams(paramNamedList);
      LMDirichletSimilarityFactory factory = new LMDirichletSimilarityFactory();
      factory.init(params);
      Similarity simLMDir = factory.getSimilarity();
      searcher.setSimilarity(simLMDir);
    }

    QueryParser p = new QueryParser(Version.LUCENE_43, StatusField.TEXT.name,
        IndexStatuses.ANALYZER);
View Full Code Here
TOP

Related Classes of org.apache.lucene.search.similarities.Similarity

  • cc.twittertools.search.api.TrecSearchHandler
  • cc.twittertools.search.indexing.SearchStatuses
  • cc.twittertools.search.local.RunQueries
  • cc.twittertools.search.local.SearchStatuses
  • cc.twittertools.search.retrieval.QueryEnvironment
  • cc.twittertools.search.retrieval.TrecSearchHandler
  • cc.wikitools.lucene.WikipediaSearcher
  • org.apache.blur.manager.clusterstatus.ZookeeperClusterStatus
  • org.apache.blur.manager.indexserver.DistributedIndexServer
  • org.apache.lucene.index.memory.MemoryIndex$MemoryIndexReader
    • 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.