Statistical Dialogue Management using Intention Dependency Graph
Statistical Dialogue Management using Intention Dependency Graph
复制标题
DOI：
--
复制
发表时间：
2013-10
期刊：
影响因子：
--
通讯作者：
Koichiro Yoshino;Shinji Watanabe;Jonathan Le Roux;J. Hershey
中科院分区：
文献类型：
--
作者：
Koichiro Yoshino;Shinji Watanabe;Jonathan Le Roux;J. Hershey
提出了一种基于部分可观察马尔可夫决策过程（POMDP）框架的有向意图依赖图（IDG）的统计对话管理方法。该模型中的转移概率涉及从意图的层次图中获得的信息。通过这种方式，我们将传统基于规则的系统的确定性图结构与统计对话框架结合起来。IDG还提供了对用户模拟模型的合理约束，在学习POMDP和对话评估中的策略功能时使用该模型。因此，该方法将传统的对话管理器转换为统计对话管理器，该管理器利用任务域知识而不带注释的对话数据。
We present a method of statistical dialogue management using a directed intention dependency graph (IDG) in a partially observable Markov decision process (POMDP) framework. The transition probabilities in this model involve information derived from a hierarchical graph of intentions. In this way, we combine the deterministic graph structure of a conventional rule-based system with a statistical dialogue framework. The IDG also provides a reasonable constraint on a user simulation model, which is used when learning a policy function in POMDP and dialogue evaluation. Thus, this method converts a conventional dialogue manager to a statistical dialogue manager that utilizes task domain knowledge without annotated dialogue data.