Improving the Generalizability of Collaborative Dialogue Analysis With Multi-Feature Embeddings

Ayesha Enayet, Gita Sukthankar


Abstract
Conflict prediction in communication is integral to the design of virtual agents that support successful teamwork by providing timely assistance. The aim of our research is to analyze discourse to predict collaboration success. Unfortunately, resource scarcity is a problem that teamwork researchers commonly face since it is hard to gather a large number of training examples. To alleviate this problem, this paper introduces a multi-feature embedding (MFeEmb) that improves the generalizability of conflict prediction models trained on dialogue sequences. MFeEmb leverages textual, structural, and semantic information from the dialogues by incorporating lexical, dialogue acts, and sentiment features. The use of dialogue acts and sentiment features reduces performance loss from natural distribution shifts caused mainly by changes in vocabulary. This paper demonstrates the performance of MFeEmb on domain adaptation problems in which the model is trained on discourse from one task domain and applied to predict team performance in a different domain. The generalizability of MFeEmb is quantified using the similarity measure proposed by Bontonou et al. (2021). Our results show that MFeEmb serves as an excellent domain-agnostic representation for meta-pretraining a few-shot model on collaborative multiparty dialogues.
Anthology ID:
2023.eacl-main.258
Volume:
Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics
Month:
May
Year:
2023
Address:
Dubrovnik, Croatia
Editors:
Andreas Vlachos, Isabelle Augenstein
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3551–3565
Language:
URL:
https://aclanthology.org/2023.eacl-main.258
DOI:
10.18653/v1/2023.eacl-main.258
Bibkey:
Cite (ACL):
Ayesha Enayet and Gita Sukthankar. 2023. Improving the Generalizability of Collaborative Dialogue Analysis With Multi-Feature Embeddings. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 3551–3565, Dubrovnik, Croatia. Association for Computational Linguistics.
Cite (Informal):
Improving the Generalizability of Collaborative Dialogue Analysis With Multi-Feature Embeddings (Enayet & Sukthankar, EACL 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.eacl-main.258.pdf
Video:
 https://aclanthology.org/2023.eacl-main.258.mp4