Sentence Compression as Tree Transduction

Abstract

This paper presents a tree-to-tree transduction method for sentence compression. Our model is based on synchronous tree substitution grammar, a formalism that allows local distortion of the tree topology and can thus naturally capture structural mismatches. We describe an algorithm for decoding in this framework and show how the model can be trained discriminatively within a large margin framework. Experimental results on sentence compression bring significant improvements over a state-of-the-art model.

References in corpus (1)

• Sentence Compression as Tree Transduction

Cited by in corpus (7)

• Sentence Compression as Tree Transduction
• Learning to Perform Local Rewriting for Combinatorial Optimization
• Neural Summarization by Extracting Sentences and Words
• Centrality-as-Relevance: Support Sets and Similarity as Geometric Proximity
• Leveraging Graph to Improve Abstractive Multi-Document Summarization
• Deep Tree Transductions - A Short Survey
• Sequence-to-Sequence Learning with Latent Neural Grammars