A BK-tree is a metric tree suggested by Walter Austin Burkhard and Robert M. Keller[1] specifically adapted to discrete metric spaces. For simplicity, consider integer discrete metric . Then, BK-tree is defined in the following way. An arbitrary element a is selected as root node. The root node may have zero or more subtrees. The k-th subtree is recursively built of all elements b such that . BK-trees can be used for approximate string matching in a dictionary.[2][example needed]

Example edit

 
An example of BK-tree

This picture depicts the BK-tree for the set   of words {"book", "books", "cake", "boo", "boon", "cook", "cake", "cape", "cart"} obtained by using the Levenshtein distance

  • each node   is labeled by a string of  ;
  • each arc   is labeled by   where   denotes the word assigned to  .

The BK-tree is built so that:

  • for all node   of the BK-tree, the weight assigned to its egress arcs are distinct;
  • for all arc   labeled by  , each descendant   of   satisfies the following equation:  :
    • Example 1: Consider the arc from "book" to "books". The distance between "book" and any word in {"books", "boo", "boon", "cook"} is equal to 1;
    • Example 2: Consider the arc from "books" to "boo". The distance between "books" and any word in {"boo", "boon", "cook"} is equal to 2.

Insertion edit

The insertion primitive is used to populate a BK-tree   according to a discrete metric  .

Input:

  •  : the BK-tree;
    •   denotes the weight assigned to an arc  ;
    •   denotes word assigned to a node  );
  •  : the discrete metric used by   (e.g. the Levenshtein distance);
  •  : the element to be inserted into  ;

Output:

  • The node of   corresponding to  

Algorithm:

  • If the   is empty:
    • Create a root node   in  
    •  
    • Return  
  • Set   to the root of  
  • While   exists:
    •  
    • If  :
      • Return  
    • Find   the child of   such that  
    • If   is not found:
      • Create the node  
      •  
      • Create the arc  
      •  
      • Return  
    •  

Lookup edit

Given a searched element  , the lookup primitive traverses the BK-tree to find the closest element of  . The key idea is to restrict the exploration of   to nodes that can only improve the best candidate found so far by taking advantage of the BK-tree organization and of the triangle inequality (cut-off criterion).

Input:

  •  : the BK-tree;
  •  : the corresponding discrete metric (e.g. the Levenshtein distance);
  •  : the searched element;
  •  : the maximum distance allowed between the best match and  , defaults to  ;

Output:

  •  : the closest element to   stored in   and according to   or   if not found;

Algorithm:

  • If   is empty:
    • Return  
  • Create   a set of nodes to process, and insert the root of   into  .
  •  
  • While  :
    • Pop an arbitrary node   from  
    •  
    • If  :
      •  
    • For each egress-arc  :
      • If  : (cut-off criterion)
        • Insert   into  .
  • Return  

Example of the lookup algorithm edit

Consider the example 8-node B-K Tree shown above and set  "cool".   is initialized to contain the root of the tree, which is subsequently popped as the first value of   with  ="book". Further   since the distance from "book" to "cool" is 2, and   as this is the best (i.e. smallest) distance found thus far. Next each outgoing arc from the root is considered in turn: the arc from "book" to "books" has weight 1, and since   is less than  , the node containing "books" is inserted into   for further processing. The next arc, from "book" to "cake," has weight 4, and since   is not less than  , the node containing "cake" is not inserted into  . Therefore, the subtree rooted at "cake" will be pruned from the search, as the word closest to "cool" cannot appear in that subtree. To see why this pruning is correct, notice that a candidate word   appearing in "cake"s subtree having distance less than 2 to "cool" would violate the triangle inequality: the triangle inequality requires that for this set of three numbers (as sides of a triangle), no two can sum to less than the third, but here the distance from "cool" to "book" (which is 2) plus the distance from "cool" to   (which is less than 2) cannot reach or exceed the distance from "book" to "cake" (which is 4). Therefore, it is safe to disregard the entire subtree rooted at "cake".

Next the node containing "books" is popped from   and now  , the distance from "cool" to "books." As  ,   remains set at 2 and the single outgoing arc from the node containing "books" is considered. Next, the node containing "boo" is popped from   and  , the distance from "cool" to "boo." This again does not improve upon  . Each outgoing arc from "boo" is now considered; the arc from "boo" to "boon" has weight 1, and since  , "boon" is added to  . Similarly, since  , "cook" is also added to  .

Finally each of the two last elements in   are considered in arbitrary order: suppose the node containing "cook" is popped first, improving   to distance 1, then the node containing "boon" is popped last, which has distance 2 from "cool" and therefore does not improve the best result. Finally, "cook" is returned as the answer   with  .

See also edit

References edit

External links edit

  • A BK-tree implementation in Common Lisp with test results and performance graphs.
  • An explanation of BK-Trees and their relationship to metric spaces [3]
  • An explanation of BK-Trees with an implementation in C# [4]
  • A BK-tree implementation in Lua [5]
  • A BK-tree implementation in Python [6]