A Prime Number Approach to Matching an XML Twig Pattern including Parent-Child Edges

Abstract

Abstract:Twig pattern matching is a core operation in XML query processing because it is how all the occurrences of a twig pattern in an XML document are found. In the past decade, many algorithms have been proposed to perform twig pattern matching. They rely on labelling schemes to determine relationships between elements corresponding to query nodes in constant time. In this paper, a new algorithm TwigStackPrime is proposed, which is an improvement to TwigStack (Bruno et al., 2002). To reduce the memory consumption and computation overhead of twig pattern matching algorithms when Parent-Child (P-C) edges are involved, TwigStackPrime efficiently filters out a tremendous number of irrelevant elements by introducing a new labelling scheme, called Child Prime Label (CPL). Extensive performance studies on various real-world and artificial datasets were conducted to demonstrate the significant improvement of CPL over the previous indexing and querying techniques. The experimental results show that the new technique has a superior performance to the previous approaches.

TwigStackPrime: A Novel Twig Join Algorithm Based on Prime Numbers

“…The key idea of their approach is based on the proposition that when visiting document elements in post-order, it can then be determined whether or not they contribute to the final result before storing them in intermediate storage which is trees of stacks to ensure linear processing. TwigList [11] replaced the complex intermediate storage proposed in T wig 2 Stack with lists (one for every query node) and pointers with simple intervals to capture structural relationships. The authors in [7] proposed a new storage scheme, level vector split which splits the list connected to its parent list with P-C edge to a number of levels bounded by the maximum depth of the XML tree.…”

“…4 and Q 2 = a[/x]/y/f, the head elements in their streams are a → a 1 , x → x 1 , y → y 1 and f → f 1 . The first call of getNext(root) inside the main algorithm will return a → a 1 because it has A-D relationship with all head elements and satisfies CPL with x and y, and its descendant y → y 1 also satisfies the child and descendant extension with respect to f. However, TwigStackPrime produces the useless path (a 1 , x 1 ) because y 2 does not have child of f-node [2].…”

“…As a result, an XML query is defined as a complex selection on elements of an XML document specified by structural information of the selected elements. Improving the efficiency of tree patterns matching is a [2,10,9,17,4] since tree patterns are the basis for querying structured tree-based data model such as XML.…”

“…Both node and graph indexing are essential to XML query processing algorithms, they play important role in providing efficient evaluation of queries with respect to CPU complexity and memory consumption overhead [22,2]. According to [23], a labelling scheme has to guarantee uniqueness and order preservation of node labels, thus the hierarchical relationships between a pair of nodes can be determined efficiently.…”

“…Illustration to the suboptimal processing of TwigStack. TwigStackPrime[2] Input: TPQ Q 1 while ¬end(getRoot(Q)) do 2 qact = getNext(getRoot(Q)) // see Algorithm3 if ¬ isRoot(q) then 4 cleanStack(getElement(qact), parent(qact)) 5 if isRoot(q) ∨ ¬ empty(S parent(q act ) ) then 6 cleanStack(getElement(qact),qact) Function cleanStack(Query node qact,Query node q): 14 // pop any element in Sq which is not the ancestor of getElement(qact) 15 while ¬empty(Sq) ∧ getEnd(top(Sq)) < getStart(getElement(qact)) do 16 pop(Sq) 17 Function moveToStack(Query node q): 18 // p is a pointer to the top parent stack if q is the root p is null 19 // p = top(S parent(q) ) 20 push(Cq, p) to Sq 21 Function end(Query node q):…”