Our project provides several types of results :

Detailed Guidelines concerning each level of segmentations

Segmented data according to annotation levels for German and French :

  • a French pilot corpus of ten excerpts of 10 minutes, aligned in tokens and syllables
  • a German pilot corpus of twelve transcripts of ten interaction types of ten minutes, aligned with the audio signal

automatic Tools to segment oral data for French chunks, German pauses and German syntax

Data exploration in an analytic perspective