Revisiting Conversation Discourse for Dialogue Disentanglement
Open Access
- 30 November 2024
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Information Systems
- Vol. 43 (1) , 1-34
Abstract
Dialogue disentanglement aims to detach the chronologically ordered utterances into several independent sessions. Conversation utterances are essentially organized and described by the underlying discourse, and thus dialogue disentanglement requires the full understanding and harnessing of the intrinsic discourse attribute. In this article, we propose enhancing dialogue disentanglement by taking full advantage of the dialogue discourse characteristics. First of all, in feature encoding stage , we construct the heterogeneous graph representations to model the various dialogue-specific discourse structural features, including the static speaker-role structures (i.e., speaker-utterance and speaker-mentioning structure) and the dynamic contextual structures (i.e., the utterance-distance and partial-replying structure). We then develop a structure-aware framework to integrate the rich structural features for better modeling the conversational semantic context. Second, in model learning stage , we perform optimization with a hierarchical ranking loss mechanism, which groups dialogue utterances into different discourse levels and carries training covering pairwise and session-wise levels hierarchically. Third, in inference stage , we devise an easy-first decoding algorithm, which performs utterance pairing under the easy-to-hard manner with a global context, breaking the constraint of traditional sequential decoding order. On two benchmark datasets, our overall system achieves new state-of-the-art performances on all evaluations. In-depth analyses further demonstrate the efficacy of each proposed idea and also reveal how our methods help advance the task. Our work has great potential to facilitate broader multi-party multi-thread dialogue applications.Keywords
All Related Versions
Funding Information
- National Natural Science Foundation of China (62176187)
- Social Science Fund of the Ministry of Education (20YJA740062)
- CCF-Kuaishou Large Model Explorer Fund
- CCF-Baidu Open Fund
- China Scholarship Council
This publication has 53 references indexed in Scilit:
- Memory Graph with Message Rehearsal for Multi-Turn Dialogue Generation2022
- Conversation Disentanglement with Bi-Level Contrastive Learning2022
- Modeling Compositionality with Dependency Graph for Dialogue Generation2022
- HSAN: A Hierarchical Self-Attention Network for Multi-Turn Dialogue Generation2021
- Multi-Turn Dialogue Reading Comprehension With Pivot Turns and KnowledgeIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2021
- Neural Multimodal Belief Tracker with Adaptive Attention for Dialogue Systems2019
- Disentangling ChatComputational Linguistics, 2010
- Comparing Clusterings by the Variation of Information2003
- Long Short-Term MemoryNeural Computation, 1997
- Comparing partitionsJournal of Classification, 1985