Dialogue Session Segmentation by Embedding-Enhanced TextTiling

Yiping Song, Lili Mou, Rui Yan, Li Yi, Zinan Zhu, Xiaohua Hu, Ming Zhang


In human-computer conversation systems, the context of a user-issued utterance is particularly important because it provides useful background information of the conversation. However, it is unwise to track all previous utterances in the current session as not all of them are equally important. In this paper, we address the problem of session segmentation. We propose an embedding-enhanced TextTiling approach, inspired by the observation that conversation utterances are highly noisy, and that word embeddings provide a robust way of capturing semantics. Experimental results show that our approach achieves better performance than the TextTiling, MMD approaches.


DOI: 10.21437/Interspeech.2016-1234

Cite as

Song, Y., Mou, L., Yan, R., Yi, L., Zhu, Z., Hu, X., Zhang, M. (2016) Dialogue Session Segmentation by Embedding-Enhanced TextTiling. Proc. Interspeech 2016, 2706-2710.

Bibtex
@inproceedings{Song+2016,
author={Yiping Song and Lili Mou and Rui Yan and Li Yi and Zinan Zhu and Xiaohua Hu and Ming Zhang},
title={Dialogue Session Segmentation by Embedding-Enhanced TextTiling},
year=2016,
booktitle={Interspeech 2016},
doi={10.21437/Interspeech.2016-1234},
url={http://dx.doi.org/10.21437/Interspeech.2016-1234},
pages={2706--2710}
}