Package elki.index.tree.metrical.mtreevariants.strategies.split

Class MLBDistSplit<E extends MTreeEntry,​N extends AbstractMTreeNode<?,​N,​E>>

  • Type Parameters:
    E - the type of MTreeEntry used in the M-Tree
    N - the type of AbstractMTreeNode used in the M-Tree
    All Implemented Interfaces:
    MTreeSplit<E,​N>
    @Priority(200)
@Reference(authors="P. Ciaccia, M. Patella, P. Zezula",
           title="M-tree: An Efficient Access Method for Similarity Search in Metric Spaces",
           booktitle="Proc. Int. Conf. Very Large Data Bases (VLDB\'97)",
           url="http://www.vldb.org/conf/1997/P426.PDF",
           bibkey="DBLP:conf/vldb/CiacciaPZ97")
public class MLBDistSplit<E extends MTreeEntry,​N extends AbstractMTreeNode<?,​N,​E>>
extends AbstractMTreeSplit<E,​N>
    Encapsulates the required methods for a split of a node in an M-Tree. The routing objects are chosen according to the MLBDIST strategy.

    The benefit of this strategy is that it works with precomputed distances from the parent, while most other strategies would require O(n²) distance computations. So if construction time is critical, this is a good choice.

    Reference:

    P. Ciaccia, M. Patella, P. Zezula
    M-tree: An Efficient Access Method for Similarity Search in Metric Spaces
    Proc. Int. Conf. Very Large Data Bases (VLDB'97)

    Since:
    0.6.0
    Author:
    Elke Achtert

    • Constructor Detail

      • MLBDistSplit

        public MLBDistSplit​(DistributionStrategy distributor)
        Constructor.
        Parameters:
        distributor - Distribution strategy

    • Method Detail

      • split

        public Assignments<E> split​(AbstractMTree<?,​N,​E,​?> tree,
                            N node)
        Selects the second object of the specified node to be promoted and stored into the parent node and partitions the entries according to the M_LB_DIST strategy.

        This strategy considers all possible pairs of objects and chooses the pair of objects for which the distance is maximum.

        Parameters:
        tree - Tree to use
        node - the node to be split
        Returns:
        the assignments of this split