Knowledge Representation and Inductive Learning with XML
Xiaobing Wu · Web Intelligence · 2004
This paper presents a novel knowledge representation method and learning system for XML documents. The traditional machine learning methods which use attribute-value languages are not suitable for representing XML documents due to their complex structures. In this paper, we propose a decision-tree algorithm for XML learning, which is based on a rich representation language for structured data and driven by precision/recall heuristic.