Package org.apache.lucene.search.similarities

Examples of org.apache.lucene.search.similarities.Similarity
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
    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, IndexStatuses.StatusField.TEXT.name, IndexStatuses.ANALYZER);
    Query query = p.parse(queryText);
View Full Code Here
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
    public NumericDocValues getNormValues(String field) {
      FieldInfo fieldInfo = fieldInfos.get(field);
      if (fieldInfo == null || fieldInfo.omitsNorms())
        return null;
      NumericDocValues norms = cachedNormValues;
      Similarity sim = getSimilarity();
      if (!field.equals(cachedFieldName) || sim != cachedSimilarity) { // not cached?
        Info info = getInfo(field);
        int numTokens = info != null ? info.numTokens : 0;
        int numOverlapTokens = info != null ? info.numOverlapTokens : 0;
        float boost = info != null ? info.getBoost() : 1.0f;
        FieldInvertState invertState = new FieldInvertState(field, 0, numTokens, numOverlapTokens, 0, boost);
        long value = sim.computeNorm(invertState);
        norms = new MemoryIndexNormDocValues(value);
        // cache it for future reuse
        cachedNormValues = norms;
        cachedFieldName = field;
        cachedSimilarity = sim;
View Full Code Here
95
96
97
98
99
100
101
102
103
104
105
  private ConnectedMoreLikeThisQueryBuilder.INPUT_TYPE inputType;
  private TermQuery findById;

  public MoreLikeThisBuilder( DocumentBuilderIndexedEntity documentBuilder, SearchFactoryImplementor searchFactory ) {
    this.documentBuilder = documentBuilder;
    Similarity configuredSimilarity = searchFactory.getIndexBindings().get( documentBuilder.getBeanClass() ).getSimilarity();
    if ( configuredSimilarity instanceof TFIDFSimilarity ) {
      this.similarity = (TFIDFSimilarity) configuredSimilarity;
    }
    else {
      throw log.requireTFIDFSimilarity( documentBuilder.getBeanClass() );
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
  public void testFloatNorms() throws IOException {

    Directory dir = newDirectory();
    IndexWriterConfig config = newIndexWriterConfig(TEST_VERSION_CURRENT,
        new MockAnalyzer(random()));
    Similarity provider = new MySimProvider();
    config.setSimilarity(provider);
    RandomIndexWriter writer = new RandomIndexWriter(random(), dir, config);
    final LineFileDocs docs = new LineFileDocs(random());
    int num = atLeast(100);
    for (int i = 0; i < num; i++) {
View Full Code Here
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
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
    public NumericDocValues getNormValues(String field) {
      FieldInfo fieldInfo = fieldInfos.get(field);
      if (fieldInfo == null || fieldInfo.omitsNorms())
        return null;
      NumericDocValues norms = cachedNormValues;
      Similarity sim = getSimilarity();
      if (!field.equals(cachedFieldName) || sim != cachedSimilarity) { // not cached?
        Info info = getInfo(field);
        int numTokens = info != null ? info.numTokens : 0;
        int numOverlapTokens = info != null ? info.numOverlapTokens : 0;
        float boost = info != null ? info.getBoost() : 1.0f;
        FieldInvertState invertState = new FieldInvertState(field, 0, numTokens, numOverlapTokens, 0, boost);
        long value = sim.computeNorm(invertState);
        norms = new MemoryIndexNormDocValues(value);
        // cache it for future reuse
        cachedNormValues = norms;
        cachedFieldName = field;
        cachedSimilarity = sim;
View Full Code Here
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
    List<AtomicReaderContext> leaves = topReaderContext.leaves();
    int subIndex = ReaderUtil.subIndex(11, leaves);
    for (int i = 0, c = leaves.size(); i < c; i++) {
      final AtomicReaderContext ctx = leaves.get(i);
    
      final Similarity sim = new DefaultSimilarity() {
        @Override
        public float sloppyFreq(int distance) {
          return 0.0f;
        }
      };
 
      final Similarity oldSim = searcher.getSimilarity();
      Scorer spanScorer;
      try {
        searcher.setSimilarity(sim);
        SpanNearQuery snq = new SpanNearQuery(
                                new SpanQuery[] {
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
115
116
117
118
119
120
121
122
123
124
125
  public void buildIndex(Directory dir) throws IOException {
    Random random = random();
    IndexWriterConfig config = newIndexWriterConfig(TEST_VERSION_CURRENT,
        new MockAnalyzer(random()));
    Similarity provider = new MySimProvider();
    config.setSimilarity(provider);
    RandomIndexWriter writer = new RandomIndexWriter(random, dir, config);
    final LineFileDocs docs = new LineFileDocs(random, defaultCodecSupportsDocValues());
    int num = atLeast(100);
    for (int i = 0; i < num; i++) {
View Full Code Here
81
82
83
84
85
86
87
88
89
90
91
      SearchConfiguration cfg,
      WorkerBuildContext buildContext
  ) {
    String indexName = getIndexName( entity, cfg );
    Properties[] indexProperties = getIndexProperties( cfg, indexName );
    Similarity similarity = createSimilarity( indexName, cfg, indexProperties[0], entity, buildContext );
    boolean isDynamicSharding = isShardingDynamic( indexProperties[0], buildContext );

    IndexManager[] indexManagers = new IndexManager[0];
    if ( !isDynamicSharding ) {
      indexManagers = createIndexManagers(
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.