Research map / Reinforcement learning / Imitation learning
Task-Oriented Dialog Systems That Consider Multiple Appropriate Responses under the Same Context
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, human evaluation | diversity score: 3.12 -> 3.65 | Helped |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, human evaluation | appropriateness score: 2.50 -> 2.53 | Helped |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, human evaluation | invalid responses (%): 6.1% -> 9.9% | Hurt |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, oracle belief and generated action | combined score: 100.9 -> 102.2 | Helped |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, oracle belief and generated action | inform rate (%): 89.5 -> 89.2 | Hurt |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD · MultiWOZ, oracle belief and generated action | success rate (%): 75.8 -> 77.9 | Helped |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD with top-k sampling · MultiWOZ, 5-action generation | act number: 3.08 -> 3.43 | Helped |
| Add other observed valid actions for the same dialog state during training (ablation) | DAMD with top-p sampling · MultiWOZ, 10-action generation | slot number: 4.41 -> 6.17 | Helped |
| Generate belief state instead of providing oracle belief state (ablation) | DAMD + MADA · MultiWOZ, generated belief and generated action | combined score: 102.2 -> 85.0 | Hurt |
| Use oracle system action rather than generate the action (ablation) | DAMD + MADA · MultiWOZ, oracle belief | combined score: 102.2 -> 118.5 | Helped |
| Use top-k sampling rather than standard beam search for action generation (ablation) | DAMD · MultiWOZ, 5-action generation | act number: 2.67 -> 3.08 | Helped |
| Use top-p sampling rather than beam search for action generation (ablation) | DAMD · MultiWOZ, 10-action generation | slot number: 4.06 -> 4.41 | Helped |
Closest papers by meaning
- TA&AT: Enhancing Task-Oriented Dialog with Turn-Level Auxiliary Tasks and Action-Tree Based Scheduled Sampling (AAAI 2024 · similarity 0.90)
- MultiTalk: A Highly-Branching Dialog Testbed for Diverse Conversations (AAAI 2021 · similarity 0.88)
- Winnie: Task-Oriented Dialog System with Structure-Aware Contrastive Learning and Enhanced Policy Planning (AAAI 2024 · similarity 0.86)
- Learning to Clarify: Multi-turn Conversations with Action-Based Contrastive Self-Training (ICLR 2025 · similarity 0.86)
- From Retrieval to Generation: A Simple and Unified Generative Model for End-to-End Task-Oriented Dialogue (AAAI 2024 · similarity 0.86)
- ToolDial: Multi-turn Dialogue Generation Method for Tool-Augmented Language Models (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.