Dialog state tracking, a machine reading approach using Memory Network

Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 305–314, Valencia, Spain, April 3-7, 2017. c⃝2017 Association for Computational Linguistics Dialog state tracking, a machine reading approach using Memory Network Julien Perez Xerox Research Centre Europe Grenoble, France julien.perez@xrce.xerox.com Fei Liu ∗ The University of Melbourne Victoria, Australia fliu3@student.unimelb.edu.au Abstract In an end-to-end dialog system, the aim of dialog state tracking is to accurately estimate a compact representation of the current dialog status from a sequence of noisy observations produced by the speech recognition and the natural language un- derstanding modules. This paper intro- duces a novel method of dialog state track- ing based on the general paradigm of machine reading and proposes to solve it using an End-to-End Memory Net- work, MemN2N, a memory-enhanced neu- ral network architecture. We evaluate the proposed approach on the second Dia- log State Tracking Challenge (DSTC-2) dataset. The corpus has been converted for the occasion in order to frame the hid- den state variable inference as a question- answering task based on a sequence of ut- terances extracted from a dialog. We show that the proposed tracker gives encourag- ing results. Then, we propose to extend the DSTC-2 dataset and the definition of this dialog state task with specific reason- ing capabilities like counting, list mainte- nance, yes-no question answering and in- definite knowledge management. Finally, we present encouraging results using our proposed MemN2N based tracking model. 1 Introduction One of the core components of state-of-the-art and industrially deployed dialog systems is a dialog state tracker. Its purpose is to provide a compact representation of a dialog produced from past user inputs and system outputs which is called the di- alog state. The dialog state summarizes the infor- ∗Work carried out as an intern at XRCE mation needed to successfully maintain and finish a dialog, such as users’ goals or requests. In the simplest case of a so-called slot-filling schema, the state is composed of a predefined set of variables with a predefined domain of expression for each of them. As a matter of fact, in the recent con- text of end-to-end trainable machine learnt dialog systems, state tracking remains a central element of such architectures (Wen et al., 2016). Current models, mainly based on the principle of discrim- inative learning, tend to share three common lim- itations. First, the tracking task is perform using a fixed window of the past dialog utterances as support for decision. Second, the possible cor- relations between the set of tracked variables are not leveraged and individual trackers tend to be learnt independently. Third, the tracking task is summarized as the capability of inferring values for a predefined set of latent variables. Starting from these observations, we propose to formalize the task of state tracking as a particular instance of machine reading problem. Indeed, these formal- ization and the proposed resolution model called MemN2N (Weston et al., 2015) allow to define a tracker that is be …

Content truncated to 500 words