On the Use of Machine Learning and Syllable Information in European Portuguese Grapheme-Phone Conversion
- 356 Downloads
In this study evaluation of two self-learning methods (MBL and TBL) on European Portuguese grapheme-to-phone conversion is presented. Combinations (parallel and cascade) of the two systems were also tested. The usefulness of syllable information is also investigated. Systems with good performance were obtained both using a single self-learning method and combinations. Best performance was obtained with MBL and the parallel combination. The use of syllable information contributes to a better performance in all systems tested.
Unable to display preview. Download preview PDF.
- 1.Marchand, Y., Damper, R.: Can syllabification improve pronunciation by analogy of english? Natural Language Engineering 1(1), 1–25 (2005)Google Scholar
- 2.Reichel, U.D., Schiel, F.: Using morphology and phoneme history to improve grapheme-tophoneme conversion. In: Proc. InterSpeech Lisboa (2005)Google Scholar
- 3.Oliveira, C., Moutinho, L.C., Teixeira, A.: On European Portuguese automatic syllabification. In: Proc. InterSpeech Lisboa (2005)Google Scholar
- 4.Brill, E.: Transformation-based error-driven learning and natural language processing: a case study in part-of-speech tagging. Computational Linguistics 21, 543–565 (1995)Google Scholar
- 5.Florian, R., Ngai, G.: Fast Transformation-Based Learning (2001)Google Scholar
- 6.Daelemans, W., Zavrel, J., van der Sloot, K., van den Bosch, A.: TiMBL: Tilburg memory based learner, version 5.1, reference guide. Reference Guide ILK-0402, Tilburg University, Reference: ILK Research Group Technical Report Series no. 04-02 (2004)Google Scholar
- 7.Nascimento, F., Marques, L., Segura, L.: Português fundamental: Métodos e documentos. Technical report, INIC-CLUL, Lisboa (1987)Google Scholar