TY - GEN

T1 - A bit-parallel tree matching algorithm for patterns with horizontal VLDC's

AU - Tsuji, Hisashi

AU - Ishino, Akira

AU - Takeda, Masayuki

PY - 2005

Y1 - 2005

N2 - The tree pattern matching problem is, given two labeled trees P and T, respectively called pattern tree and target tree, to find all occurrences of P within T. Many studies have been undertaken on this problem for both the cases of ordered and unordered trees. To realize flexible matching, a kind of variable-length-don't-care's (VLDC's) have been introduced. In particular, the path-VLDC's appear in XPath, a language for addressing parts of an XML document. In this paper, we introduce horizontal VLDC's, each matches a sequence of trees whose root nodes are consecutive siblings in ordered trees. We address the tree pattern matching problem for patterns with horizontal VLDC's, In our setting, the target tree is given as a tagged sequence such as XML data stream. We present an algorithm that solves the problem in O(mn) time using O(mh) space, where m and n are the sizes of P and T, respectively, and h is the height of T. We adopt the bit-parallel technique to obtain a practically fast algorithm.

AB - The tree pattern matching problem is, given two labeled trees P and T, respectively called pattern tree and target tree, to find all occurrences of P within T. Many studies have been undertaken on this problem for both the cases of ordered and unordered trees. To realize flexible matching, a kind of variable-length-don't-care's (VLDC's) have been introduced. In particular, the path-VLDC's appear in XPath, a language for addressing parts of an XML document. In this paper, we introduce horizontal VLDC's, each matches a sequence of trees whose root nodes are consecutive siblings in ordered trees. We address the tree pattern matching problem for patterns with horizontal VLDC's, In our setting, the target tree is given as a tagged sequence such as XML data stream. We present an algorithm that solves the problem in O(mn) time using O(mh) space, where m and n are the sizes of P and T, respectively, and h is the height of T. We adopt the bit-parallel technique to obtain a practically fast algorithm.

UR - http://www.scopus.com/inward/record.url?scp=33646720695&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33646720695&partnerID=8YFLogxK

U2 - 10.1007/11575832_43

DO - 10.1007/11575832_43

M3 - Conference contribution

AN - SCOPUS:33646720695

SN - 3540297405

SN - 9783540297406

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 388

EP - 398

BT - String Processing and Information Retrieval - 12th International Conference, SPIRE 2005, Proceedings

T2 - 12th International Conference on String Processing and Information Retrieval, SPIRE 2005

Y2 - 2 November 2005 through 4 November 2005

ER -