Frequent Subtree Based Indexing

Abstract

Frequent patterns in a database can provide information on how to build efficient indexing structures for the databases. A new indexing method, called discriminative, frequent subtree_based indexing, first generates all frequent subtrees, and select discriminative subtrees among them as indexing features, then translates subtrees in the feature set into sequences, and holds them in a prefix tree. Frequent substructure explore the intrinsic characteristics of the data and are relatively stable to database updates. Discriminative, frequent subtree_based indexing can improve dramatically the performance of subtree search.