Paper: Encoding Tree Pair-Based Graphs in Learning Algorithms: The Textual Entailment Recognition Case

ACL ID W08-2004
Title Encoding Tree Pair-Based Graphs in Learning Algorithms: The Textual Entailment Recognition Case
Venue Coling 2008: Proceedings of the workshop on Speech Processing for Safety Critical Translation and Pervasive Applications
Session
Year 2008
Authors

In this paper, we provide a statistical ma- chine learning representation of textual en- tailment via syntactic graphs constituted by tree pairs. We show that the natural way of representing the syntactic relations be- tween text and hypothesis consists in the huge feature space of all possible syntac- tic tree fragment pairs, which can only be managed using kernel methods. Experi- ments with Support Vector Machines and our new kernels for paired trees show the validity of our interpretation.