Elamur Research⌕
☰
home / papers / 2009 / Grammar Engineering for CCG using Ant and XSLT

Grammar Engineering for CCG using Ant and XSLT

Paper · 2009 · Scott Martin, Rajakrishnan Rajkumar and Michael White

Summary

The authors challenge the widespread notion that corpus conversion and grammar extraction are done once and never revisited. They propose treating these steps as part of a process that is continually refined. The input is the standard CCGbank. The system adds information to it that is absent from the original: Propbank roles, head lexicalization for case-marking prepositions, derivation restructuring for the analysis of punctuation, named entities, and lemmas. First, the corpus is converted to XML using JavaCC. Then a chain of XSLT transformations, orchestrated by Apache Ant, produces the converted corpus and a grammar in OpenCCG format. The corpus and the grammar remain consistent with each other, which is useful for machine learning, in particular for realization ranking. Configuration is factored out into Ant tasks, separate from the code, so experiments do not require recompilation and different strategies can be tested more quickly. The resulting grammar achieved state-of-the-art BLEU for surface realization on Section 23 of CCGbank. The authors emphasize that improvements to the engineering process itself mattered as much as better statistical models.

Key points

  • Input: CCGbank (Hockenmaier and Steedman, 2007). Output: a converted corpus in XML and an OpenCCG grammar.
  • The corpus is augmented with Propbank roles, head lexicalization for case-marking prepositions, derivation restructuring for punctuation, named entity annotation, and lemmatization.
  • Tools: Ant's built-in javacc task (which runs JavaCC to convert to XML), two custom Ant tasks, convert and extract, and the Ant data types FileSet and FileList.
  • Result: state-of-the-art BLEU for surface realization on Section 23 of CCGbank. The extracted text gives no specific figures.
  • Future plans: increase the number of single-rooted LFs and integrate the system with OpenCCG.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓The Ohio State UniversityColumbus · United States90 works · 19,112 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 18,614 worksArtificial Intelligence · 17,965 worksArts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 1Research subjects and languages the work is about.Syntax & grammar formalisms · 221 works

Cite

Cite as Scott Martin, Rajakrishnan Rajkumar and Michael White. 2009. Grammar Engineering for CCG using Ant and XSLT. Proceedings of the Workshop on Software Engineering, Testing, and Quality Assurance for Natural Language Processing (SETQA-NLP 2009).

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{martin2009grammar,
  title = {{Grammar Engineering for CCG using Ant and XSLT}},
  author = {Scott Martin, Rajakrishnan Rajkumar and Michael White},
  booktitle = {Proceedings of the Workshop on Software Engineering, Testing, and Quality Assurance for Natural Language Processing (SETQA-NLP 2009)},
  year = {2009},
  doi = {10.3115/1621947.1621955},
  url = {https://doi.org/10.3115/1621947.1621955}
}

References

6 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Stephen Boxwell and Michael White. 2008. Projecting Propbank roles onto the CCGbank. In Proc. LREC-08. find
  2. Michael Burke, Aoife Cahill, Mairead Mccarthy, Ruth O’Donovan, Josef Genabith, and Andy Way. 2004. Evaluating automatic LFG F-structure annotation for the Penn-II treebank. Research on Language and Computation, 2:523–547, December. find
  3. Dominic Espinosa, Michael White, and Dennis Mehay. 2008. Hypertagging: Supertagging for surface realization with CCG. In Proc. ACL-08: HLT. find
  4. Julia Hockenmaier and Mark Steedman. 2007. CCGbank: A Corpus of CCG Derivations and Dependency Structures Extracted from the Penn Treebank. Computational Linguistics, 33(3):355–396.
  5. Michael White and Rajakrishnan Rajkumar. 2008. A more precise analysis of punctuation for broad-coverage surface realization with CCG. In Proc. of the Workshop on Grammar Engineering Across Frameworks (GEAF08).
  6. Michael White, Rajakrishnan Rajkumar, and Scott Martin. 2007. Towards broad coverage surface realization with CCG. In Proc. of the Workshop on Using Corpora for NLG: Language Generation and Machine Translation (UC-NLG+MT).

Updated

This page in Markdown → https://md.research.elamur.ai/free/en/papers/martin-2009-grammar-engineering-ccg-ant-xslt/