Learning Dialog Systems 2017


Janus Recognition Toolkit (JRTk) has been used by the Interactive System Lab in many projects for speech recognition, such as BABEL.  The Babel Program (IARPA) is developing agile and robust speech recognition technology that can be rapidly applied to any human language in order to provide effective search capability for analysts to efficiently process massive amounts of real-world recorded speech.



See also:

100 Best Deep Learning Videos | 100 Best GitHub: Deep Learning

SimpleDS: A Simple Deep Reinforcement Learning Dialogue System
H Cuayáhuitl – Dialogues with Social Robots, 2017 – Springer
Abstract This article presents SimpleDS, a simple and publicly available dialogue system trained with deep reinforcement learning. In contrast to previous reinforcement learning dialogue systems, this system avoids manual feature engineering by performing action

Active learning for example-based dialog systems
T Hiraoka, G Neubig, K Yoshino, T Toda… – Dialogues with Social …, 2017 – Springer
Abstract While example-based dialog is a popular option for the construction of dialog systems, creating example bases for a specific task or domain requires significant human effort. To reduce this human effort, in this paper, we propose an active learning framework to

Scaling up deep reinforcement learning for multi-domain dialogue systems
H Cuayáhuitl, S Yu, A Williamson… – Neural Networks (IJCNN …, 2017 – ieeexplore.ieee.org
Standard deep reinforcement learning methods such as Deep Q-Networks (DQN) for multiple tasks (domains) face scalability problems due to large search spaces. This paper proposes a three-stage method for multi-domain dialogue policy learning-termed NDQN,

Investigation of Language Understanding Impact for Reinforcement Learning Based Dialogue Systems
X Li, YN Chen, L Li, J Gao, A Celikyilmaz – arXiv preprint arXiv:1703.07055, 2017 – arxiv.org
Abstract: Language understanding is a key component in a spoken dialogue system. In this paper, we investigate how the language understanding module influences the dialogue system performance by conducting a series of systematic experiments on a task-oriented

Effect of a Humanoid’s Active Role during Learning with Embodied Dialogue System
M Kerzel, HG Ng, S Griffiths, S Wermter – Proceedings of the Workshop – researchgate.net
Abstract—When humanoid robots learn complex sensorimotor abilities from interaction with the environment, often a human experimenter is required. For a social companion robot, it is desirable that the learning can also be assisted by non-expert users. To achieve this aim, we

Composite Task-Completion Dialogue System via Hierarchical Deep Reinforcement Learning
B Peng, X Li, L Li, J Gao, A Celikyilmaz, S Lee… – arXiv preprint arXiv …, 2017 – arxiv.org
Abstract: In a composite-domain task-completion dialogue system, a conversation agent often switches among multiple sub-domains before it successfully completes the task. Given such a scenario, a standard deep reinforcement learning based dialogue agent may suffer

Reward-Balancing for Statistical Spoken Dialogue Systems using Multi-objective Reinforcement Learning
S Ultes, P Budzianowski, I Casanueva, N Mrkši?… – arXiv preprint arXiv …, 2017 – arxiv.org
Abstract: Reinforcement learning is widely used for dialogue policy optimization where the reward function often consists of more than one component, eg, the dialogue success and the dialogue length. In this work, we propose a structured method for finding a good balance

BBQ-Networks: Efficient Exploration in Deep Reinforcement Learning for Task-Oriented Dialogue Systems
Z Lipton, X Li, J Gao, L Li, F Ahmed, L Deng – arXiv preprint arXiv …, 2017 – arxiv.org
Abstract: We present a new algorithm that significantly improves the efficiency of exploration for deep Q-learning agents in dialogue systems. Our agents explore via Thompson sampling, drawing Monte Carlo samples from a Bayes-by-Backprop neural network. Our

VOILA: An Optimised Dialogue System for Interactively Learning Visually-Grounded Word Meanings (Demonstration System)
Y Yu, A Eshghi, O Lemon – Proceedings of the 18th Annual SIGdial …, 2017 – aclweb.org
Abstract We present VOILA: an optimised, multimodal dialogue agent for interactive learning of visually grounded word meanings from a human user. VOILA is:(1) able to learn new visual categories interactively from users from scratch;(2) trained on real human-human

Statistical Spoken Dialogue Systems and the Challenges for Machine Learning
S Young – Proceedings of the Tenth ACM International …, 2017 – dl.acm.org
Abstract This talk will review the principal components of a spoken dialogue system and then discuss the opportunities for applying machine learning for building robust high performance open-domain systems. The talk will be illustrated by recent work at Cambridge

Online Learning of Attributed Bi-Automata for Dialogue Management in Spoken Dialogue Systems
M Serras, MI Torres, A Del Pozo – Iberian Conference on Pattern …, 2017 – Springer
Abstract Online learning of dialogue managers is a desirable but often costly property to obtain. Probabilistic Finite State Bi-Automata (PFSBA) have shown to provide a flexible and adaptive framework to achieve this goal. In this paper, an Attributed PFSBA (A-PSFBA) is

Learning Generative End-to-end Dialog Systems with Knowledge
T Zhao – 2017 – cs.cmu.edu
Abstract Dialog systems are intelligent agents that can converse with human in natural language and facilitate human. Traditional dialog systems follow a modular approach and often have trouble expanding to new or more complex domains, which hinder the

Dialog System & Technology Challenge 6 Overview of Track 1-End-to-End Goal-Oriented Dialog learning
J Perez, YL Boureau, A Bordes – workshop.colips.org
Abstract End-to-end dialog learning is an important research subject in the domain of conversational systems. The primary task consists in learning a dialog policy from transactional dialogs of a given domain. In this context, usable datasets are needed to

A chatbot for a dialogue-based second language learning system
JX Huang, KS Lee, OW Kwon… – CALL in a climate of …, 2017 – books.google.com
… 2. GenieTutor–a task-oriented dialogue system for second-language learning We developed GenieTutor, a DB-CALL system for English learners in Korea several years ago … Task-oriented spoken dialog system for second- language learning …

Domain Complexity and Policy Learning in Task-oriented Dialogue Systems
A Papangelis, S Ultes, Y Stylianou – uni-ulm.de
Abstract In the present paper, we conduct a comparative evaluation of a multitude of information-seeking domains, using two well-known but fundamentally different algorithms for policy learning: GP-SARSA and DQN. Our goal is to gain an understanding of how the

Deep Learning for Acoustic Addressee Detection in Spoken Dialogue Systems
A Pugachev, O Akhtiamov, A Karpov… – Conference on Artificial …, 2017 – Springer
Abstract The addressee detection problem arises in real spoken dialogue systems (SDSs) which are supposed to distinguish the speech addressed to them from the speech addressed to real humans. In this work, several modalities were analyzed, and acoustic data

Deep Learning for Dialogue Systems
YN Chen, A Celikyilmaz, D Hakkani-Tür – Proceedings of ACL 2017 …, 2017 – aclweb.org
Abstract In the past decade, goal-oriented spoken dialogue systems have been the most prominent component in today’s virtual personal assistants. The classic dialogue systems have rather complex and/or modular pipelines. The advance of deep learning technologies

Learning concepts through conversations in spoken dialogue systems
R Jia, L Heck, D Hakkani-Tür… – Acoustics, Speech and …, 2017 – ieeexplore.ieee.org
Spoken dialogue systems must be able to recover gracefully from unexpected user inputs. In many cases, these unexpected utterances may be within the scope of the system, but include previously unseen phrases that the system cannot interpret. In this work, we

Sample efficient deep reinforcement learning for dialogue systems with large action spaces
G Weisz – pdfs.semanticscholar.org
Abstract In Statistical Dialogue Systems, we aim to deploy Artificial Intelligence to build automated dialogue agents that can converse with humans. A part of this effort is the policy optimisation task, which attempts to find a policy describing how to respond to humans, in

Proposal of an Intelligent Tutoring System for Procedural Learning with Context-aware Dialogue
J PALADINESab, J RAMÍREZ – icce2017.canterbury.ac.nz
… Procedural Learning with Context-aware Dialogue … Beetle II (Dzikovska, Steinhauser, Farrow, Moore, & Campbell, 2014) implements an approach based on task-oriented dialogue systems, and uses an ontology to represent domain knowledge …