This paper presents a machine learning based approach to incremental dialogue act classification with a focus on the recognition of communicative functions associated with dialogue segments in a multidimensional space, as defined in the ISO 24617-2 dialogue act annotation standard. The main goal is to establish the nature of an increment whose processing will result in a reliable overall system performance. We explore scenarios where increments are tokens or syntactically, semantically or prosodically motivated chunks. Combing local classification with meta-classifiers at a late fusion decision level we obtained state-of-the-art classification performance. Experiments were carried out on manually corrected transcriptions and on potentially erroneous ASR output. Chunk-based classification yields better results on the manual transcriptions, whereas token-based classification shows a more robust performance on the ASR output. It is also demonstrated that layered hierarchical and cascade training procedures result in better classification performance than the single-layered approach based on a joint classification predicting complex class labels.
Cite as: Ebhotemhen, E., Petukhova, V., Klakow, D. (2017) Incremental Dialogue Act Recognition: Token- vs Chunk-Based Classification. Proc. Interspeech 2017, 889-893, doi: 10.21437/Interspeech.2017-738
@inproceedings{ebhotemhen17_interspeech,
author={Eustace Ebhotemhen and Volha Petukhova and Dietrich Klakow},
title={{Incremental Dialogue Act Recognition: Token- vs Chunk-Based Classification}},
year=2017,
booktitle={Proc. Interspeech 2017},
pages={889--893},
doi={10.21437/Interspeech.2017-738}
}