Efficient discovery of frequent unordered trees

Siegfried Nijssen, Joost N. Kok · 2003

Abstract. Recently, an algorithm called Freqt was introduced which enumerates all frequent induced subtrees in an ordered data tree. We propose a new algorithm for mining unordered frequent induced subtrees. We show that the complexity of enumerating unordered trees is not higher than the complexity of enumerating ordered trees; a strategy for determining the frequency of unordered trees is introduced. 1

Read the paper · More papers on PaperTik