Sylli - The SSP Syllabifier
Sylli is a universal syllabifier. Developed for Italian, it can easily be adapted to any language that is claimed to respect the SSP. Sylli divides timit, strings, files and directories into syllables and provides other useful function for syllable analysis.
Even though Sylli comes out from a long tradition, it differs from other syllabifiers in many features:
- it just does not try to obtain the syllabification out of a black box, nor by implementing an arbitrary number of rules or transformations. Sylli can be used to explore the principle it is based on and was designed with particular attention to phonological theories.
- It is modular. You can edit the sonority scale, test the ouptut, add your own transducers, even as external program, and costumise the output, with no changes to the core structure of the Syllabification Algorithm.
- Tests demonstrate that it is 100% accurate for Italian even in resyllabification and connected speech context!
- It is easy to use and to test. Sylli comes with a GUI which allows to edit and manage its modules, to choose syllabification options, and to run a demo on chosen syllabification clusters to test.
- A command-line version can be used for scripting and advanced usage, and as a python library, which make it 100% compatible with other NLP tools such as NLTK. The code is organised in a way that reflects the model proposed. The code is also simple and richly commented.
- Last but not least it is a free, even for commercial use, cross-platform open-source software.
- 30 August 2011
- Download the handout of the talk given at Interspeech 2011.
- 27 June 2011
- A talk on Sylli will be given at Interspeech 2011, the largest international conference on speech science and technology. Get the article.
- 3 May 2011
- You can now try a basic version of Sylli online.
- 20 March 2011
- Sylli 0.9.8 has finally been released.
Sylli is a universal syllabifier.
Developed for Italian, it can easily be
adapted to any language that is claimed to respect the SSP.
Sylli divides timit, strings, files and directories into syllables
and provides other useful function for syllable analysis.
Be ready, 0.9.8 has brought major changes:
- new commented sonority
- changes to the syllabification algorithm
- code structure renewed
- command-line sylli improved
- new web-site
- new sphinxed documentation
- sonority.txt changed
- testing in *nix
- test unit added
- add clean screen button
- ajax demo online
- document algorithm and program structure
- add phonetic transcription module
- add blankspace #/+ option
- make sonority utf8
- : are converted to CC
- stdout should expand
- re-integration with CLIPS/nltk
- add to pypi and packaging (deb, rpm, osx)
- The complete updated list can be found here