Spoken Dialogue Systems (SDS) are systems which have the ability to interact with human beings using natural language as the medium of interaction. A dialogue policy plays a crucial role in determining the functioning of the dialogue management module. Handcrafting the dialogue policy is not always an option, considering the complexity of the dialogue task and the stochastic behavior of users. In recent years approaches based on Reinforcement Learning (RL) for policy optimization in dialogue management have been proved to be an efficient approach for dialogue policy optimization. Yet most of the conventional RL algorithms are data intensive and demand techniques such as user simulation. Doing so, additional modeling errors are likely to occur. This paper explores the possibility of using a set of approximate dynamic programming algorithms for policy optimization in SDS. Moreover, these algorithms are combined to a method for learning a sparse representation of the value function. Experimental results show that these algorithms when applied to dialogue management optimization are particularly sample efficient , since they learn from few hundreds of dialogue examples. These algorithms learn in an off-policy manner, meaning that they can learn optimal policies with dialogue examples generated with a quite simple strategy. Thus they can learn good dialogue policies directly from data , avoiding user modeling errors.

Reinforcement learning is now an acknowledged approach for optimizing the interaction strategy of spoken dialogue systems. If the first considered algorithms were quite basic (like SARSA), recent works concentrated on more sophisticated methods. More attention has been paid to off-policy learning, dealing with the exploration-exploitation dilemma, sample efficiency or handling non-stationarity. New algorithms have been proposed to address these issues and have been applied to dialogue management. However, each algorithm often solves a single issue at a time, while dialogue systems exhibit all the problems at once. In this paper, we propose to apply the Kalman Temporal Differences (KTD) framework to the problem of dialogue strategy optimization so as to address all these issues in a comprehensive manner with a single framework. Our claims are illustrated by experiments led on two real-world goal-oriented dialogue management frameworks, DIPPER and HIS.Index Terms-Dialogue management, reinforcement learning, spoken dialogue system.

We demonstrate the Task Completion Platform (TCP); a multi-domain dialogue platform that can host and execute large numbers of goal-orientated dialogue tasks. The platform features a task configuration language, TaskForm, that allows the definition of each individual task to be decoupled from the overarching dialogue policy used by the platform to complete those tasks. This separation allows for simple and rapid authoring of new tasks, while dialogue policy and platform functionality evolve independent of the tasks. The current platform includes machine learnt models that provide contextual slot carry-over, flexible item selection, and task selection/switching. Any new task immediately gains the benefit of these pieces of built-in platform functionality. The platform is used to power many of the multi-turn dialogues supported by the Cortana personal assistant.

No abstract