p-subsequential transducers are efficient finite-state transducers with p final outputs used in a variety of applications. Not all transducers admit equivalent p-subsequential transducers however. We briefly describe an existing generalized determinization algorithm for p-subsequential transducers and give the first characterization of p-subsequentiable transducers, transducers that admit equivalent p-subsequential transducers. Our characterization shows the existence of an efficient algorithm for testing p-subsequentiability. We have fully implemented the generalized determinization algorithm and the algorithm for testing p-subsequentiability. We report experimental results showing that these algorithms are practical in large-vocabulary speech recognition applications. The theoretical formulation of our results is the equivalence of the following three properties for finite-state transducers: determinizability in the sense of the generalized algorithm, p-subsequentiability, and the twins property. © Springer-Verlag Berlin Heidelberg 2003.
CITATION STYLE
Allauzen, C., & Mohri, M. (2003). p-Subsequentiable transducers. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2608, 24–34. https://doi.org/10.1007/3-540-44977-9_2
Mendeley helps you to discover research relevant for your work.