What is Pɛtɾʊs?

Pɛtɾʊs [1] is a an online automatic phonetic transcription system for Brazilian Portuguese [2]. For example, Pɛtɾʊs automatically converts a sequence of letters like <descrédito> into a sequence of phones [ʤiskɾɛʤɪtʊ]. In order to represent the pronunciation of a word, the IPA (International Phonetic Alphabet) [3], a latin-based alphabetical system for phonetic notation, is utilized. Beyond phonetic transcription, Pɛtɾʊs offers more information to users like: stressed syllable identification, syllabification and grammatical class (part-of-speech) annotation. Also, it transcribes hetereophonic homographs (words with the same spelling, but different pronunciation), generating both forms as output. If you would like to transcribe a single word, just go to the Try it! section. In the case that you wish to transcribe a list of words, you can go to the Upload file section.

What's syllabification?

The syllabic division consists in the identification and delimitation of the syllables that constitute each word. To perform this task automatically by Pɛtɾʊs, we adopted a rule-based algorithms described in Silva (2011) [4]. The syllabification algorithms proposed by Silva (2011, p. 39) were developed on the basis of the Brazilian Portuguese spelling, but predominating phonology of words. Thus, the words "obstrução" (occlusion), "arredondar" (to round) and "assado" (roasted) are divided into syllables as follows, respectively: (o.bs.tru.ção), (a.rre.don.dar) and (a.ssa.do).

References

1. SERRANI, Vanessa Marquiafável. Ambiente web de suporte à transcrição fonética automática de lemas em verbetes de dicionários do português do Brasil. Tese (Doutorado em Estudos Linguísticos) - Instituto de Biociências, Letras e Ciências Exatas, Universidade Estadual Paulista Julio de Mesquita Filho. São José do Rio Preto-SP, p. 202. 2015.

2. Marquiafavel, V.; Bokan, A. and Zavaglia, C. (2014). "PETRUS: A rule-based grapheme-to-phone converter for Brazilian Portuguese". In: J. Baptista et al. (Eds.): PROPOR 2014, LNAI 8776, Springer, Heidelberg (2014).

3. Cristófaro-Silva, T. (2000). "Fonética e fonologia dos português: roteiro de estudos e guia de exercícios". 3a ed., São Paulo: Contexto.

4. Cagliari, L. (2009). "Elementos de fonética do português brasileiro". São Paulo: Paulistana.

5. Silva, D. (2011). "Algoritmos de processamento da linguagem e síntese de voz com emoções aplicados a um conversor texto-fala baseado em HMM". Tese de Doutorado. Programa de Pós-Graduação em Engenharia Elétrica, COPPE, Universidade Federal do Rio de Janeiro, RJ, 2011.

About Pɛtɾʊs

The automatic phonetic transcription called Pɛtɾʊs is a project carried out in Instituto de Biociências, Letras e Ciências Exatas da UNESP (IBILCE), by the doctoral student in Linguistic Vanessa Marquiafavel Serrani, under the supervision of Profa. Dra. Claudia Zavaglia. Pɛtɾʊs is dedicated to the automatic phonetic transcription of simple words fromBrazilian Portuguese, having as reference São Paulo State standard pronunciation.

In this project, we thank Alessandro Bokan, master student from USP, for his participation and commitment.

Download

You can to download the source code! 

 Download from GitH