Last updated by wjarek on 2011-05-05. Originally submitted by wjarek on 2010-02-28.
The Department of Contemporary English Language (DoCELu) invites you to a Phon&Phon meeting:
Michał Jankowski
Automatic grapheme-to-phoneme transition for Polish revisited
Tuesday, 02 March 2010, 18:30, Room 601A
Polish linguists studying the phonology of English and/or other languages often refer to the phonology of their native language for comparison and such comparisons work best if one has access to a representative body of data in computer-readable form. The necessity to compile a comprehensive corpus of Polish phonemic transcription to be made available to my colleagues in the School of English was the initial incentive to develop a system for the automatic grapheme-to-phoneme transition for Polish. Another incentive was to see if a program written over 15 years ago to handle simple symbol transliterations could be used to do the job.
During this Phon&Phon presentation I will briefly describe various approaches to grapheme-phoneme translation and the projects that my transliteration software has been used in in the past. Further, I will show the simple mechanism behind the program and the rules that make the transliteration software change Polish orthography to phonemic transcription. I will also identify some limitations of the program. To conclude I will outline possible future projects that might make use of the techniques involved in context-sensitive symbol translation.
Submitted by J Weckwerth
Phon&Phon home page
