Class BM25Similarity


  • public class BM25Similarity
    extends Similarity
    BM25 Similarity. Introduced in Stephen E. Robertson, Steve Walker, Susan Jones, Micheline Hancock-Beaulieu, and Mike Gatford. Okapi at TREC-3. In Proceedings of the Third Text REtrieval Conference (TREC 1994). Gaithersburg, USA, November 1994.
    • Field Detail

      • discountOverlaps

        protected boolean discountOverlaps
        True if overlap tokens (tokens with a position of increment of zero) are discounted from the document's length.
    • Constructor Detail

      • BM25Similarity

        public BM25Similarity​(float k1,
                              float b)
        BM25 with the supplied parameter values.
        Parameters:
        k1 - Controls non-linear term frequency normalization (saturation).
        b - Controls to what degree document length normalizes tf values.
      • BM25Similarity

        public BM25Similarity()
        BM25 with these default values:
        • k1 = 1.2,
        • b = 0.75.
    • Method Detail

      • idf

        protected float idf​(long docFreq,
                            long numDocs)
        Implemented as log(1 + (numDocs - docFreq + 0.5)/(docFreq + 0.5)).
      • sloppyFreq

        protected float sloppyFreq​(int distance)
        Implemented as 1 / (distance + 1).
      • scorePayload

        protected float scorePayload​(int doc,
                                     int start,
                                     int end,
                                     BytesRef payload)
        The default implementation returns 1
      • avgFieldLength

        protected float avgFieldLength​(CollectionStatistics collectionStats)
        The default implementation computes the average as sumTotalTermFreq / maxDoc, or returns 1 if the index does not store sumTotalTermFreq (Lucene 3.x indexes or any field that omits frequency information).
      • encodeNormValue

        protected byte encodeNormValue​(float boost,
                                       int fieldLength)
        The default implementation encodes boost / sqrt(length) with SmallFloat.floatToByte315(float). This is compatible with Lucene's default implementation. If you change this, then you should change decodeNormValue(byte) to match.
      • setDiscountOverlaps

        public void setDiscountOverlaps​(boolean v)
        Sets whether overlap tokens (Tokens with 0 position increment) are ignored when computing norm. By default this is true, meaning overlap tokens do not count when computing norms.
      • getDiscountOverlaps

        public boolean getDiscountOverlaps()
        Returns true if overlap tokens are discounted from the document's length.
        See Also:
        setDiscountOverlaps(boolean)
      • computeNorm

        public final long computeNorm​(FieldInvertState state)
        Description copied from class: Similarity
        Computes the normalization value for a field, given the accumulated state of term processing for this field (see FieldInvertState).

        Matches in longer fields are less precise, so implementations of this method usually set smaller values when state.getLength() is large, and larger values when state.getLength() is small.

        Specified by:
        computeNorm in class Similarity
        Parameters:
        state - current processing state for this field
        Returns:
        computed norm value
      • idfExplain

        public Explanation idfExplain​(CollectionStatistics collectionStats,
                                      TermStatistics[] termStats)
        Computes a score factor for a phrase.

        The default implementation sums the idf factor for each term in the phrase.

        Parameters:
        collectionStats - collection-level statistics
        termStats - term-level statistics for the terms in the phrase
        Returns:
        an Explain object that includes both an idf score factor for the phrase and an explanation for each term.
      • computeWeight

        public final Similarity.SimWeight computeWeight​(float queryBoost,
                                                        CollectionStatistics collectionStats,
                                                        TermStatistics... termStats)
        Description copied from class: Similarity
        Compute any collection-level weight (e.g. IDF, average document length, etc) needed for scoring a query.
        Specified by:
        computeWeight in class Similarity
        Parameters:
        queryBoost - the query-time boost.
        collectionStats - collection-level statistics, such as the number of tokens in the collection.
        termStats - term-level statistics, such as the document frequency of a term across the collection.
        Returns:
        SimWeight object with the information this Similarity needs to score a query.