Greek verbs and user friendliness in the speech recognition and the speech production module of dialog systems for the broad public

  • Authors:
  • Christina Alexandris;Ioanna Malagardi

  • Affiliations:
  • National and Kapodistrian University of Athens, Greece;Department of Informatics & Telecommunications, Educational & Language Technology Laboratory, National and Kapodistrian University of Athens

  • Venue:
  • UAHCI'11 Proceedings of the 6th international conference on Universal access in human-computer interaction: applications and services - Volume Part IV
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

It has been observed that for the Greek language, in Service- Oriented Dialog Systems targeted towards the broad public, verbs display particular features to be considered both in the System's Speech Recognition (ASR) and Speech Synthesis (or Speech Production) Module. Additionally, the function of verbs, both in respect to their role in the semantic content of the spoken utterance and in respect to their prosodic features in the spoken utterance, is observed to be directly linked to the step and the related Speech Act in the dialog structure. From a prosodic aspect, it is observed that, in spoken input, "Multitasking" verbs do not receive prosodic emphasis, whereas in spoken output, prosodic emphasis is given on words signalizing the User-System Relationship.