INTERSPEECH 2013 14thAnnual Conference of the International Speech Communication Association Lyon, France August 25-29, 2013

Failure Transitions for Joint n-Gram Models and G2P Conversion

Josef R. Novak, Nobuaki Minematsu, Keikichi Hirose

University of Tokyo, Japan

This work investigates two related issues in the area of WFST-based G2P conversion. The first is the impact that the approach utilized to convert a target word to an equivalent finite-state machine has on downstream decoding efficiency. The second issue considered is the impact that the approach utilized to represent the joint n-gram model via the WFST framework has on the speed and accuracy of the system. In the latter case two novel algorithms are proposed, which extend the work from [1] to enable the use of failure-transitions with joint n-gram models. All solutions presented in this work are available as part of the open-source, BSD-licensed Phonetisaurus G2P toolkit [2].

References

1. C. Allauzen, M. Mohri, and B. Roark, “Generalized algorithms for constructing statistical language models,” in Proceedings of the 41st Annual Meeting on Association for Computational Linguistics, vol.1, 2003, 40–47.
2. J. Novak, “Phonetisaurus g2p,” 2012. [Online]. Available: http://code.google.com/p/phonetisaurus

Full Paper

Bibliographic reference.  Novak, Josef R. / Minematsu, Nobuaki / Hirose, Keikichi (2013): "Failure transitions for joint n-gram models and G2p conversion", In INTERSPEECH-2013, 1821-1825.