ISCA Archive Interspeech 2012
ISCA Archive Interspeech 2012

Perma and Balloon: tools for string alignment and text processing

Uwe D. Reichel

Two research tools available as webservices are presented in this paper: PermA, a general-purpose string aligner which can for example be used for grapheme-to-phoneme and phoneme-to-phoneme alignment, and Balloon, a text processing toolkit for German and English providing components for part-of-speech tagging, morphological analyses, and grapheme-to-phoneme conversion including syllabification and wordstress assignment. In this paper the general architectures of these tools are introduced with a focus on recent enhancements concerning the alignment cost function derivation and word stress assignment.

Index Terms: alignment, grapheme-to-phoneme conversion, part-of-speech tagging, morphology, word-stress assignment, tools