Elamur Research⌕
☰
home / papers / 1994 / Simplified English grammar and style correction in an MT framework: The LRE SECC Project

Simplified English grammar and style correction in an MT framework: The LRE SECC Project

Summary

The paper describes the SECC project of the LRE-2 programme, which ran from November 1993 to May 1996. Its goal was to build a tool for technical writers that checks whether documents comply with the rules of Simplified English (SE) and corrects violations automatically where possible. The central idea is to treat checking and correction as translation from English into SE, that is, as a separate language pair in the transfer-based MT system Metal. The usual "analysis–transfer–generation" cycle becomes "analysis–diagnosis–correction". The existing English lexicon and analysis grammar are reused. Transfer detects errors and replaces disallowed words, while generation produces error messages and, where possible, a corrected sentence. The author distinguishes two kinds of diagnosis. Weak diagnosis flags a possible error and only gives a hint. Strong diagnosis flags a definite error that can be corrected automatically. The interfaces are also described, in particular integration into Interleaf through the Metal system's API. There are no quantitative results yet, as this is a report on work in progress. Still, it shows that tools built for machine translation can be applied to other NLP tasks.

Key points

  • The consortium includes Siemens Nixdorf, Cap Gemini, Alcatel Bell, the University of Leuven and Sietec. The total effort is about 13 person-years.
  • The SE variant has three parts. COLEX contains about 1500 allowed words plus about 500 more replacements (for example, rapid/quick → fast). COTECH is the technical terminology of telephony. COGRAM consists of 150 rules in four groups, about two thirds of which can be implemented computationally.
  • Testing used an Alcatel Bell corpus of 2000 sentences. The Metal English lexicon contains over 50 000 base forms.
  • Example correction: the possessive form of an inanimate noun is rewritten with an of-construction (module's function → function of the module).
  • Future plans: support for Flemish, French and German writers (relaxed rule conditions, fallback rules, and the user can switch this function off), checking at the text level rather than only the sentence level, and an interactive mode in Interleaf using SGML.

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

Fields: Linguistics & languages · Machine translation

Cite

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

@inproceedings{adriaens1994simplified,
  title = {{Simplified English grammar and style correction in an MT framework: The LRE SECC Project}},
  author = {Adriaens, Geert},
  booktitle = {Proceedings of Translating and the Computer 16},
  year = {1994},
  doi = {10.1108/eb051383},
  url = {https://doi.org/10.1108/eb051383}
}

References

12 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Adriaens, G. and Schreurs D. "From Cogram to Alcogram: towards a Controlled English Grammar Checker," Proceedings of COLING92,595-601,1992. find
  2. Adriaens, G., Deprez F., De Braekeleer G., and Depoortere B. "Sisyphus: a toolkit for MT testing and evaluating," Unpublished paper, Siemens Nixdorf Software Centre Liege, 1993. For clarity's sake: both Dutch people (i.e. people living in the Netherlands) and Flemish people (i.e. people living in the northern part of Belgium) have Dutch as their language. The technical writers in the consortium are Flemish. find
  3. Alonso, J. A. "The Spanish Grammar Checker," TWB (Translator's Workbench) Final Report, 1992. find
  4. Humphreys, L. "The Simplified English Lexicon," Proceedings of EURALEX92,353-364,1992. find
  5. Pulman, S.G. and Rayner, M. "Computer Processable Controlled Language," SRI International Cambridge Computer Science Research Centre, 1994. find
  6. Ravin, Y. "Grammar Errors and Style Weaknesses in a Text-Critiquing System," K. Jensen, J.E. Heidorn and S.D. Richardson (eds), Natural Language Processing: The PLNLP Approach. Kluwer Academic Publishers, Boston, 65-76,1993. ↗
  7. Schreurs, D. and Adriaens, G. "Controlled English (CE): From COGRAM to ALCOGRAM," In P. O'Brian Holt and N. Williams (eds), Computers and Writing, State of the Art. Intellect Books Oxford and Kluwer Academic Publishers, Boston, 206-221,1992. ↗
  8. Steen, G.J. van der, and Dijenborg, B. "Linguistic Engineering: Tools and Products," Proceedings of the Twente Workshop on Language Technology 2, University of Twente, 1992. find
  9. Thurmair, G. "Parsing for Grammar and Style Checking," Proceedings of COLING90,356-370,1990. find
  10. Thurmair, G. "Verification of Controlled Grammars," Esprit Translator's Workbench Report. Also in Translator's Workbench Final Report (abbreviated version), 128-132,1992. find
  11. Wojcik, R.H., Hoard, J.E. and Holzhauser, K.C. "The Boeing Simplified English Checker," Proceedings of the International Conference, Human Machine Interaction and Artificial Intelligence in Aeronautics and Space. Toulouse, Centre d'Etudes et de Recherches de Toulouse, 43-57, 1990. find
  12. Wojcik, R. H., Harrison, P. and Bremer, J. "Using Bracketed Parses to Evaluate a Grammar Checking Application," Proceedings of ACL93, 38-45, 1993. find

This page in Markdown → https://md.research.elamur.ai/free/en/papers/adriaens-1994-simplified-english-grammar-style-correction/