Realising the Translation of Utterances into Meanings by Petri Net Transducers

Abstract:

In this paper, we illustrate by a small case study how the translation of utterances into meanings within a hierarchical cognitive dynamic speech signal processing system can be realised by Petri net transducers (PNTs). PNTs are a natural generalisation of finite state transducers (FSTs) for the transla- tion of partial languages consisting of partial words (with a partial order on their symbols) instead of (linear) words (having a total order on their symbols). For the considered case study we extend previous definitions of PNTs by weights and composition operations. We use bisemirings for the set of weights of a PNT.


Year: 2013
In session: Kognitive Systeme
Pages: 103 to 110