0
Learning Personalized End-to-End Goal-Oriented Dialog
本文是来自于AAAI 2019关于在任务型对话中引入个性化的文章，主要通过用户profile建模来实现个性化回答和语义消歧。
Tag: Task Oriented Dialogue
0
Discriminative Deep Dyna-Q Robust Planning for Dialogue Policy Learning
本文是Deep Dyna-Q Integrating Planning for Task-Completion Dialogue Policy Learning 团队的续作，主要解决的是原始DDQ模型对world model生成的simulated dialogues质量好坏的严重依赖，通过引入一个区分真实对话和模拟对话的判别器，进而提高DDQ模型的鲁棒性和有效性。paper linkcode link
0
Deep Dyna-Q Integrating Planning for Task-Completion Dialogue Policy Learning
本文提出了一种新的通过与真实用户交互来学习对话策略的方法Deep Dyna-Q，与之前的工作相比，只需要少量的真实对话数据，通过world model对用户建模，结合model-free和model-based，该方法能够高效的学习对话策略。
0
Frame
A brief introduction to Maluuba’s Frames dataset. It is designed to help drive research that enables truly conversational agents that can support decision-making in complex settings. The dataset contains natural and complex dialogues with users considering different options, comparing packages, and progressively building rich descriptions through conversation.
0
Latent Intention Dialogue Models
论文提出了一种隐意图对话模型（Latent Intention Dialogue Model, LIDM），通过离散的隐变量来学习对话意图，这些隐变量可以看作引导对话生成的动作决策，提高基于手工构建的状态-动作集传统强化学习模型所生成对话的多样性。