Learn More
The lack of suitable training and testing data is currently a major roadblock in applying machine-learning techniques to dialogue management. Stochastic modelling of real users has been suggested as a solution to this problem, but to date few of the proposed models have been quantitatively evaluated on real data. Indeed , there are no established criteria(More)
Recent work in the area of probabilistic user simulation for training statistical dialogue managers has investigated a new agenda-based user model and presented preliminary experiments with a handcrafted model parameter set. Training the model on dialogue data is an important next step, but non-trivial since the user agenda states are not observable in data(More)
Over the past decade, a variety of user models have been proposed for user simulation-based reinforcement-learning of dialogue strategies. However, the strategies learned with these models are rarely evaluated in actual user trials and it remains unclear how the choice of user model affects the quality of the learned strategy. In particular , the degree to(More)
1996 The initial impetus for the current popularity of statistical methods in computational linguistics was provided in large part by the papers on part-of-speech tagging by Church [20], DeRose [25], and Garside [34]. In contradiction to common wisdom, these taggers showed that it was indeed possible to carve part-of-speech disambiguation out of the(More)
This work shows how a dialogue model can be represented as a factored Partially Observable Markov Decision Process (POMDP). The fac-tored representation has several benefits, such as enabling more nuanced reward functions to be specified. Although our dialogue model is significantly larger than past work using POMDPs, experiments on a small testbed problem(More)
Partially Observable Markov Decision Processes (POMDPs) have been shown to be a promising framework for dialog management in spoken dialog systems. However, to date, POMDPs have been limited to artificially small tasks. In this work, we present a novel method called a " Summary POMDP " for scaling slot-filling POMDP-based dialog managers to cope with tasks(More)
This work shows how a spoken dialogue system can be represented as a Partially Observable Markov Decision Process (POMDP) with composite observations consisting of discrete elements representing dialogue acts and continuous components representing confidence scores. Using a testbed simulated dialogue management problem and recently developed optimisation(More)
Although partially observable Markov decision processes (POMDPs) have shown great promise as a framework for dialog management in spoken dialog systems, important scalability issues remain. This paper tackles the problem of scaling slot-filling POMDP-based dialog managers to many slots with a novel technique called composite point-based value iteration(More)