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 paper, 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 pair-wise 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.

Learning to Disentangle Interleaved Conversational Threads with a Siamese Hierarchical Network and Similarity Ranking

Differentiating Interpreting Types: Connecting Complex Networks to Cognitive Complexity

Chinese Writing of Deaf or Hard-of-hearing Students and Normal-hearing Peers from Complex Network Approach

Learning The Structures Of Online Asynchronous Conversations

POSLAN: Disentangling Chat with Positional and Language encoded Post Embeddings

Revisiting Conversation Discourse for Dialogue Disentanglement

Hierarchical Dialogue Understanding with Special Tokens and Turn-level Attention

Interpreting Hierarchical Linguistic Interactions in DNNs

Hierarchical Summarization for Longform Spoken Dialog

Structural Modeling for Dialogue Disentanglement

CM-Net: A Novel Collaborative Memory Network for Spoken Language Understanding

Channel-Aware Decoupling Network for Multiturn Dialog Comprehension.

CLUCDD: CONTRASTIVE DIALOGUE DISENTANGLEMENT VIA CLUSTERING

Structural Characterization for Dialogue Disentanglement

Learning Semantic Textual Similarity from Conversations

UniMC: A Unified Framework for Long-Term Memory Conversation via Relevance Representation Learning

HFD: Hierarchical feature decoupling for SQL generation from text

Dance with Labels: Dual-Heterogeneous Label Graph Interaction for Multi-intent Spoken Language Understanding.

Improving Text Semantic Similarity Modeling through a 3D Siamese Network

Natural Language Processing and Chinese Computing - 9th CCF International Conference, NLPCC 2020, Zhengzhou, China, October 14-18, 2020, Proceedings, Part I.