Algorithms for Finding a Most Similar Subforest

摘要

Given an ordered labeled forest F ("the target forest") and an ordered labeled forest G ("the pattern forest"), the most similar subforest problem is to find a subforest F′ of F such that the distance between F′ and G is minimum over all possible F′. This problem generalizes several well-studied problems which have important applications in locating patterns in hierarchical structures such as RNA molecules' secondary structures and XML documents. In this paper, we present efficient algorithms for the most similar subforest problem with forest edit distance for three types of subforests: simple substructures, sibling substructures, and closed subforests.