Papers by Thomas Ruprecht

3 papers
Supertagging-based Parsing with Linear Context-free Rewriting Systems (2021.naacl-main)

Copied to clipboard

Challenge: a new supertagging-based parser for linear context-free rewriting systems is developed for discontinuous constituents . discontinuous constituencies span non-contiguous sets of positions in a sentence, and can be modelled by CFG .
Approach: They propose a supertagging-based parser for linear context-free rewriting systems . they propose an efficient procedure which induces a lexical LCFRS from any discontinuous treebank .
Outcome: The proposed method outperforms previous LCFRS-based parsers in accuracy and speed by a wide margin.
Implementation of a Chomsky-Schützenberger n-best parser for weighted multiple context-free grammars (N19-1)

Copied to clipboard

Challenge: Constituent parsing has been studied extensively in the last decades.
Approach: They propose to decompose a language into a regular language, a homomorphism, and a bracket language to divide the parsing problem into simpler subproblems.
Outcome: The proposed approach is comparable to state-of-the-art grammar-based parsers.
Improving the Extraction of Supertags for Constituency Parsing with Linear Context-Free Rewriting Systems (2022.findings-emnlp)

Copied to clipboard

Challenge: a new approach to parsing discontinuous constituency structures uses supertags to improve parsability . traditional approaches use grammar formalisms to model hierarchies of noncontiguous phrases . but supertags are still useful for analyzing these grammars and parsers .
Approach: They propose to reformulate and parameterize extraction process for LCFRS supertags to improve parsing quality.
Outcome: The proposed method improves the quality and speed of parsing with supertags over the previous method.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations