Deterministic natural language generation from meaning representations for machine translation

What it is about

The paper argues for a deterministic generator that starts from meaning representations, for use as the target side of a machine translation pipeline. The input is a Penman-style representation derived from Davidsonian predicate formulas. The generator turns it into a parse tree in the style of a historical English treebank, and the words are the yield of that tree. Generation proceeds through a fixed sequence of tree transformations. The main idea is that the way the information is arranged in the meaning representation can decide which grammatical construction is chosen, such as a passive or a cleft. The paper shows many constructions with their decision rules. It makes no statistical claim, and the choice between constructions is fixed by the arrangement, not ranked.

Ideas

• I1. An information-arrangement feature on meta tokens. Let the reader mark which meta tokens are given and which are new, from the verse context, and let the encoder pick a construction table from that mark. This is one of the function-word and boundary features of open problem 4.
• I2. A logged sequence of transformations. Write the encoder of tools/translate as an ordered list of tree transforms, and record which one fired on each token. Count the firings per author. This gives the explanation the project asks for, and the counts show which transforms each author uses.
• I3. Author-specific constructions as deltas. Find verses where two Bible translations render the same meta tokens with different constructions, such as active and passive. Count how many such verses there are per pair. A construction that differs in many verses is a candidate for an author delta, which needs at least three fixed tokens to be kept.

For Mova

The paper fits the encoder side of tools/translate (profile.rs) and the ordered rule chain that Mova already uses for the target stack. The arrangement feature is a small addition to meta.rs. The deterministic sequence of transforms is the design to copy, with the log of firings as the explanation.

Topics: Machine translation

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/w16-0601

@inproceedings{butler2016deterministic,
  title = {{Deterministic natural language generation from meaning representations for machine translation}},
  author = {Butler, Alastair},
  booktitle = {Proceedings of the 2nd Workshop on Semantics-Driven Machine Translation (SedMT 2016)},
  year = {2016},
  doi = {10.18653/v1/w16-0601},
  url = {https://doi.org/10.18653/v1/w16-0601}
}

References

23 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Hiyan Alshawi. 1992. The Core Language Engine. Cambridge, Mass.: MIT Press. find
2. Yoav Artzi, Kenton Lee, and Luke Zettlemoyer. 2015. Broad-coverage CCG semantic parsing with AMR. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, page 1699–1710. Lisbon, Portugal. find
3. L. Banarescu, C. Bonial, S. Cai, M. Georgescu, K. Griffitt, U. Hermjakob, K. Knight, P. Koehn, M. Palmer, and N. Schneider. 2013. Abstract Meaning Representation for Sembanking. In Proceedings of the 7th Linguistic Annotation Workshop and Interoperability with Discourse, pages 178–186. find our card
4. Ann Bies, Mark Ferguson, Karen Katz, and Robert MacIntyre. 1995. Bracketing guidelines for Treebank II style Penn Treebank project. Tech. Rep. MS-CIS-95-06, LINC LAB 281, University of Pennsylvania Computer and Information Science Department. find
5. Johan Bos. 2008. Wide-coverage semantic analysis with Boxer. In J. Bos and R. Delmonte, eds., Semantics in Text Processing. STEP 2008 Conference Proceedings, Research in Computational Semantics, pages 277–286. College Publications. find
6. Alastair Butler. 2015a. Linguistic Expressions and Semantic Processing: A Practical Approach. Heidelberg: Springer-Verlag. find
7. Alastair Butler. 2015b. Round-trips with meaning stopovers. In Proceedings of the 1st Workshop on Semantics-Driven Statistical Machine Translation, pages 1–10. Beijing, China: Association for Computational Linguistics. find
8. Noam Chomsky. 1970. Remarks on nominalization. In R. Jacobs and P. Rosenbaum, eds., Readings in English Transformational Grammar, pages 184–221. Waltham, MA: Ginn. find
9. Ann Copestake, Dan Flickinger, Carl Pollard, and Ivan A. Sag. 2006. Minimala Recursion Semantics: an introduction. Research on Language and Computation 3(4):281–332. find
10. Donald Davidson. 1967. The logical form of action sentences. In N. Rescher, ed., The Logic of Decision and Action. Pittsburgh: University of Pittsburgh Press. Reprinted in: D. Davidson, 1980. Essays on Actions and Events. Claredon Press, Oxford, pages 105–122. find
11. Jeffrey Flanigan, Sam Thomson, Jaime Carbonell, Chris Dyer, and Noah A. Smith. 2014. A discriminative graph-based parser for the abstract meaning representation. In Proc. of the ACL 2014. find
12. Kevin Humphreys, Mike Calcagno, and David Weise. 2001. Reusing a statistical language model for generation. In Proceedings of the 8th European workshop on Natural Language Generation - Volume 8, pages 1–6. Stroudsburg, PA: Association for Computational Linguistics. find
13. Anthony Kroch, Beatrice Santorini, and Ariel Diertani. 2010. The Penn-Helsinki Parsed Corpus of Modern British English (PPCMBE). Department of Linguistics, University of Pennsylvania. CD-ROM, second edition, (http://www.ling.upenn.edu/hist-corpora). ↗
14. Fred Landman. 2000. Events and Plurality: The Jerusalem Lectures. Dordrecht: Kluwer Academic Publishers. find
15. Irene Langkilde and Kevin Knight. 1998. Generation that exploits corpus-based statistical knowledge. In Proceedings of the ACL/COLING-98. Montreal, Quebéc. find
16. Roger Levy and Galen Andrew. 2006. Tregex and tsurgeon: tools for querying and manipulating tree data structure. In 5th International conference on Language Resources and Evaluation. find
17. Christian Matthiessen and John A Bateman. 1991. Text generation and systemic-functional linguistics: experiences from English and Japanese. Pinter Publishers. find
18. Koji Mineshima, Pascual Martínez-Gómez, Yusuke Miyao, and Daisuke Bekki. 2015. Higher-order logical inference with compositional semantics. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP 2015), pages 2055–2061. Lisbon, Portugal. find
19. Terence Parsons. 1990. Events in the Semantics of English. Cambridge: MIT Press. find
20. Richard Pito. 1994. tgrepdoc - documentation for tgrep. find
21. A. Purtee and Lenhart K. Schubert. 2012. TTT: A tree transduction language for syntactic and semantic processing. In EACL 2012 Workshop on Applications of Tree Automata Techniques in Natural Language Processing (ATANLP 2012). Avignon, France. find
22. Michael Pust, Ulf Hermjakob, Kevin Knight, Daniel Marcu, and Jonathan May. 2015. Parsing english into abstract meaning representation using syntax-based machine translation. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1143–1154. Lisbon, Portugal. find
23. Beth Randall. 2009. CorpusSearch 2 Users Guide. (http://corpussearch.sourceforge.net/CS-manual/Contents.html). ↗

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/butler-2016-deterministic-natural-language-generation-meaning/