Robust, applied morphological generation

What it is about

Natural language generation systems often need a separate component that turns a lemma, a part of speech and an inflection type into a word form. The authors build such a generator for English on finite-state techniques. It is fast and robust, and it has been used in a prototype system that simplifies English newspaper text. The paper also discusses the morphological and orthographic problems that appear when the generator meets unrestricted text. To check the coverage of their morphology, the authors also test a morphological analyser over the British National Corpus. Against the CELEX lexical database the type accuracy is 99.97%, and the token accuracy is 99.98% over 14,825,661 relevant tokens, which is about two errors per ten thousand words.

Ideas

I1. en/src/morph.rs: a generator that takes a lemma, a part of speech and an inflection tag and returns the form, built from the lemma table data/lemmas.tsv and the form table data/forms.tsv. Measure the form accuracy on the forms of EWT test, and report the errors by inflection type.
I2. Use the paper's round-trip check: generate every form from its lemma and tag, analyse it back, and count the mismatches. Run it on the in-code dictionary and report the count per run, twice, to see that it is stable.
I3. Handle the orthographic changes (doubling, e-deletion, y to ie) in 06. Spelling rules: inflection, derivation, compounds and national standards as explicit rules, and print the rule that fired for each generated form, so that a wrong form is traceable.

For Mova

The code is en/src/morph.rs, with the command en morph-eval. The chapter is 06. Spelling rules: inflection, derivation, compounds and national standards, with 13. Verbs for verb inflection. CELEX is a licensed lexicon, so use it as a test only. Measure first on EWT test: form accuracy from lemma and tag, against the current lemma accuracy of 99.3 to 99.5%.

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.3115/1118253.1118281

@inproceedings{minnen2000robust,
  title = {{Robust, applied morphological generation}},
  author = {Minnen, Guido and Carroll, John and Pearce, Darren},
  booktitle = {INLG'2000 Proceedings of the First International Conference on Natural Language Generation},
  year = {2000},
  doi = {10.3115/1118253.1118281},
  url = {https://doi.org/10.3115/1118253.1118281}
}

References

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

of the European Chapter of the Association for Computational Linguistics, pages 87-96, Utrecht, The Netherlands. find
Yvonne Canning and John Tait. 1999. Syntactic simplification of newspaper text for aphasic readers. In Proceedings of the ACM SIGIR Workshop on Customised Information Delivery, Berkeley, CA, USA. find
John Carroll, Guido Minnen, Darren Pearce, Yvonne Canning, Siobhan Devlin, and John Tait. 1999. Simplifying English text for language impaired readers. In Pwceedings of the 9th Conference of th, e European Chapter of the Association for Computational Ling.uistics (EACL), Bergen, Norway. find
Hamish Cunningham, Yorick Wilks, and Robert Gaizauskas. 1996. GATE--a GenerM Architecture for Text Engineering. In Proceedings of the 16th Conference on Computational Linguistics, Copenhagen, Denmark. find
Siobhan Devlin and John Tait. 1998. The use of a psychotinguistic database in the simplification of text for aphasic readers. In (Nerbonne. 1998). find
Michael Elhadad and Jacques Robin. 1996. find
An Mehryar Mohri. 1996. On some applications of overview of SU-KGE:..A ~eusable~,.eomprehen- ....... :fmittee-:sta, t e -automata,.-.t heeory.,.~tox.:natu.ea, l~lam.-sive syntactic realization component. Tech-guage processing. Natural Language Engi-nical Report 96-03, Dept of Mathematics and neering… find
Databases. Lecture Notes. CSLI Publications, Stanford, USA. Roger Evans and Gerald Gazdar. 1996. DATR: a language for lexical knowledge representa-Richard Power, Donia Scott, and Roger Evans. tion. Computational Linguistics, 22. 1998. What You See Is What You Meant: direct knowledge editing~with nat… find
Paul Procter. 1995. Cambridge International Lauri Karttunen, Jean-Pierre Chanod, Gregory Dictionary of English. find
Cambridge University Grefenstette, and Anne Schiller. 1996. Regu-Press. lar expressions for language engineering. Nat-Geoffrey Pullum and Arnold Zwicky. In prepa-ural Language Engineering, 2(4):305-329. ration. Licensing of prosodic features by Lauri Karttunen. 1994. Constructing lexical syntactic r… find
Chicago, Illinois, January 1997. Available at Kimmo Koskenniemi. 1983. Two-level model \< http://www.lsadc.org/web2/99modabform.ht for morphological analysis. In 8th Interna-Philip Quinlan. 1992. The Oxford Psycholin-tional Joint Conference on Artificial Intelli-guistic Database. Oxford University Pr… ↗
Graeme Ritchie, Graham Russell, Alan Black, John Levine, Tony Mason, and Doug Brown. and Stephen Pulman. 1992. find
Computational 1992. Lex ~4 Yacc. O'Reilly and Associates, morphology: practical mechanisms for the second edition. English lexicon. MIT Press. find
Mitch Marcus, Beatrice Santorini, and Geoffrey Sampson. 1995. English for the com-Mary Ann Marcinkiewicz. 1993. Build-puter. Oxford University Press. ing a large annotated corpus of English: the Stuart Shieber, Gertjan van Noord, Robert Penn Treebank. find
Computational Linguistics, Moore, and Fernando Pereira. 1990. Seman-19(2):313-330. tic head-driven generation. find
Computational Christian Matthiessen. 1984. Systemic Gram-Linguistics, 16(1):7-17. mar in computation: The Nigel case. In Pro-Lita Taylor and Gerry Knowles. 1988. Man-ceedings of the 1st Conference of the European ual of information to accompany the SEC Chapter of the Association for Computational Co… find
UK. Fellbaum, Derek Gross, Katherine Miller, and Gertjan van Noord. 1991. Morphology in Randee Tengi. 1993. Five Papers on Word-MiMo2. Manuscript, University of Utrecht, Net. Princeton University, Princeton, N.J. The Netherlands. find
Guido Minnen and John Carroll. Under review. Past and robust morphological processing tools for practical NLP applications. find
Roger Mitton. 1992. A description of a computer-usable dictionary file based on the Oxford Advanced Learner's Dictionary of Current English. Availat)le at \< ftp: / / ota.ox.ac.uk / pub / ota/ pub lic / d icts / 710 / text710.doe: \>. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/aclbulk-2000-minnen-robust-applied-morphological-generation/