Dynamic Oracles for Top-Down and In-Order Shift-Reduce Constituent\n Parsing

2018/10/25 by Daniel Fernández‐González, Fernández-González, Daniel, Carlos Gómez‐Rodríguez +1
Biochemistry, Genetics and Molecular Biology · Computer Science · #68T50 #Computation and Language (cs.CL) #FOS: Computer and information sciences #Genomics and Phylogenetic Studies #I.2.7 #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.1810.10882

openalex publication_date 2018/10/25 · openalex created_date 2022/08/02 · openalex updated_date 2026/07/28

Abstract

We introduce novel dynamic oracles for training two of the most accurate\nknown shift-reduce algorithms for constituent parsing: the top-down and\nin-order transition-based parsers. In both cases, the dynamic oracles manage to\nnotably increase their accuracy, in comparison to that obtained by performing\nclassic static training. In addition, by improving the performance of the\nstate-of-the-art in-order shift-reduce parser, we achieve the best accuracy to\ndate (92.0 F1) obtained by a fully-supervised single-model greedy shift-reduce\nconstituent parser on the WSJ benchmark.\n

Related

• Shift-Reduce Constituent Parsing with Neural Lookahead Features
• Span-Based Constituency Parsing with a Structure-Label System and Provably Optimal Dynamic Oracles
• Enriched In-Order Linearization for Faster Sequence-to-Sequence Constituent Parsing
• In-Order Transition-based Constituent Parsing
• Transition-based Parsing with Lighter Feed-Forward Networks