LINGUIST List 6.1586

Thu Nov 9 1995

Sum: Syllabification algorithm

Editor for this issue: T. Daniel Seely <dseelyemunix.emich.edu>

Directory

1. Su-Lin Wu, summary of responses to syllabification algorithm query

Summary of responses to query about syllabification algorithms---

Many thanks to:

Larry Romans		(ljrcobra.jpl.nasa.gov)
Alain Theriault		(theriaalERE.UMontreal.CA)
Mike Hammond		(hammondaruba.ccit.arizona.edu)
David M. W. Powers	(powersist.flinders.edu.au)
Walter Daelemans	(Walter.Daelemanskub.nl)
John Coleman		(jscindy.phon.ox.ac.uk)
Hugh Buckingham		(hbucksalvador.speech.lsu.edu)
Deborah D K Ruuska	(druuskancc.helsinki.fi)
Antal van den Bosc	(antalcs.rulimburg.nl)
Ton Weijters		(weijterscs.rulimburg.nl)
Markell R West		(markellafterlife.ncsc.mil)


As you might recall, I said: "I've been looking for syllabification
algorithms, but I am having a hard time finding practical, clearly
stated and complete algorithms that will provide good coverage."

Several people mentioned the relationship of hyphenation and syllabification
and directed me to hyphenation algorithms of several flavors, though I was
cautioned that morphology plays a large role in hyphenation and none in
syllabification. Liang's computer hyphenation algorithm (used in TeX) was
mentioned as well as table-lookup methods by Weijters and
connectionist/machine learning methods by Daelemans and van den Bosch.

1)	Liang, F. M.
	Word hy-phen-a-tion by com-put-er / by Franklin Mark Liang.
 Word hyphenation by computer.
	Thesis (Ph. D.)--Stanford University, 1983.

2)	Daelemans, W. and van den Bosch, A. "Generalization
	performance of backpropagation learning on a syllabification task."
	In M.F.J. Drossaers and A Nijholt (Eds.), Proceedings of TWLT3:
	Connectionism and Natural Language Processing, pp. 27-37.
	Enschede: University of Twente. 1992.
	ftp://ftp.cs.rulimburg.nl/pub/papers/antal/twlt3-92.ps.gz

3)	Van den Bosch, A., Weijters, A., Van den Herik, J., and Daelemans,
	W. "The profit of learning exceptions". In Proceedings of the
	5th Belgian-Dutch Conference on Machine Learning, BENELEARN'95,
	pp. 118-126.
	ftp://ftp.cs.rulimburg.nl/pub/papers/antal/benelearn-95.ps.gz

4)	Weijters, A. "A simple look-up procedure superior to NETtalk?".
	In Artificial Neural Networks, T. Kohonen, K. Makisara,
	O. Simula and J. Kangas (eds)
	Elsevier Science Publishers B.V. (North-Holland), 1991.

A couple of respondents mentioned sonority theory.

1)	G. N. Clements. "The role of the sonority cycle in core
	syllabification". In PAPERS IN LABORATORY PHONOLOGY I: BETWEEN THE
	GRAMMAR AND PHYSICS OF SPEECH, John Kingston and Mary F. Beckman
	(eds). Cambridge University Press, 1990.

Also recommended for reading is

1)	R. Treiman and A. Zukowski. "Toward an understanding of english
	syllabification", In Journal of Memory and Language. 1990.
	Volume 29, pages 66-85.

Hammond's syllable parser (based on optimality theory) was mentioned by
several people and is available on-line
(http://aruba.ccit.arizona.edu/~hammond). IPOX by Coleman and
associates is also available on-line and has a grammar of English syllables
(ftp://chico.phon.ox.ac.uk/pub/ipox/index.html).

Several people wrote that they were involved in current
computer syllabification efforts in various languages, such as French, Finnish
and English and offered to send me their code.

The responses to my query provided me with several viable alternatives that
I am planning to explore.

Thank you all for your help, I really appreciate all the responses.
- Su-Lin.

- ---------------------------------------------------------------------------
Su-Lin Wu
Computer Science Division
UC Berkeley and International Computer Science Institute
sulinicsi.berkeley.edu
http://www.icsi.berkeley.edu/~sulin
- ---------------------------------------------------------------------------