Summary of the paper

Title Expert-Novice Interaction: Annotation and Analysis
Authors Soumia Dermouche and Catherine Pelachaud
Abstract In this demonstration, we present the NoXi corpus of expert-novice interactions, our annotations and analysis. To analyze the data we apply HCApriori, a Temporal Sequence Mining algorithm to extract relevant behavior sequences for both expert and novice. NoXi provides over 25 hours of dyadic interactions recorded in different languages, mainly English, French, and German. The annotation tool, NOVA, developed by (Baur et al., 2015) allows annotating data using discrete and continuous schema. We use NOVA to manually annotate non-verbal behaviors (discrete annotation) and engagement levels (continuous annotation).
Topics Sequence Mining, Non-Verbal Behavior, Engagement, Virtual Agent
Full paper Expert-Novice Interaction: Annotation and Analysis
Bibtex @InProceedings{DERMOUCHE18.15,
  author = {Soumia Dermouche and Catherine Pelachaud},
  title = {Expert-Novice Interaction: Annotation and Analysis},
  booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)},
  year = {2018},
  month = {may},
  date = {7-12},
  location = {Miyazaki, Japan},
  editor = {Hanae Koiso and Patrizia Paggio},
  publisher = {European Language Resources Association (ELRA)},
  address = {Paris, France},
  isbn = {979-10-95546-16-0},
  language = {english}
  }
Powered by ELDA © 2018 ELDA/ELRA