RNN (Recurrent Neural Network) & Dialog Systems 2016


Notes:

From 2014 to 2015, the number of academic papers in Google Scholar covering recurrent neural networks and dialog systems tripled, from around 50x in 2014 to 150x in 2015. In 2016, that number doubled again to over 300x.

  • dialog state tracker
  • dialog state tracking
  • discourse relation recognition

Resources:

  • ma3hmi.cogsy.de .. multimodal analyses enabling artificial agents in human-machine interaction (ma3hmi 2016)
  • sensei-conversation.eu .. making sense of human – human conversations (sensei fp7 project)
  • specom2016.hte.hu .. international conference on speech and computer (specom 2016)

References:

Wikipedia:

See also:

100 Best Recurrent Neural Network VideosRNN (Recurrent Neural Network) & Question Answering Systems 2016


Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models.
IV Serban, A Sordoni, Y Bengio, AC Courville, J Pineau – AAAI, 2016 – aaai.org
… Our approach differs from previous work on learning dialogue systems through inter- action with humans (Young et al. 2013; Gasic et al. 2013; Cantrell et al … We experiment with the well-established recurrent neural networks (RNN) and n-gram models …

A network-based end-to-end trainable task-oriented dialogue system
TH Wen, D Vandyke, N Mrksic, M Gasic… – arXiv preprint arXiv …, 2016 – arxiv.org
… task-oriented spoken dialogue system (SDS) [4]. Current state-of-the-art belief trackers use discriminative models such as recurrent neural networks (RNN) to directly map ASR hypotheses to belief states. Although in this work we focus on text-based dialogue systems, we retain …

Multi-domain neural network language generation for spoken dialogue systems
TH Wen, M Gasic, N Mrksic… – arXiv preprint arXiv …, 2016 – arxiv.org
… In this paper, we propose a procedure to train multi-domain, Recurrent Neural Network-based (RNN) language gen- erators via multiple … Modern Spoken Dialogue Systems (SDS) are typi- cally developed according to a well-defined ontol- ogy, which provides a structured …

A latent variable recurrent neural network for discourse relation language models
Y Ji, G Haffari, J Eisenstein – arXiv preprint arXiv:1603.01913, 2016 – arxiv.org
… In contrast to recent work on continuous latent variables in recurrent neural networks (Chung et al., 2015), which require complex variational autoen- coders … in the Switchboard cor- pus (Stolcke et al., 2000), and is a key component of contemporary dialog systems (Williams and …

How NOT to evaluate your dialogue system: An empirical study of unsupervised evaluation metrics for dialogue response generation
CW Liu, R Lowe, IV Serban, M Noseworthy… – arXiv preprint arXiv …, 2016 – arxiv.org
… Dual Encoder. Next we consider the recurrent neural network (RNN) based architecture called the Dual Encoder (DE) model (Lowe et al., 2015) … To our knowledge, our applica- tion of neural network models to large-scale re- trieval in dialogue systems is novel …

Video paragraph captioning using hierarchical recurrent neural networks
H Yu, J Wang, Z Huang, Y Yang… – Proceedings of the IEEE …, 2016 – cv-foundation.org
Page 1. Video Paragraph Captioning Using Hierarchical Recurrent Neural Networks Haonan Yu1? Jiang Wang3 Zhiheng Huang2? Yi Yang3 Wei Xu3 1Purdue University 2Facebook haonanu@gmail.com zhiheng@fb.com 3Baidu Research – Institute of Deep Learning …

Sequential short-text classification with recurrent and convolutional neural networks
JY Lee, F Dernoncourt – arXiv preprint arXiv:1603.03827, 2016 – arxiv.org
… 2016. The Fourth Dialog State Tracking Chal- lenge. In Proceedings of the 7th International Work- shop on Spoken Dialogue Systems (IWSDS). [Kim2014] Yoon Kim. 2014 … 2011. Rnnlm-recurrent neural network language mod- eling toolkit. In Proc …

Recurrent neural network grammars
C Dyer, A Kuncoro, M Ballesteros, NA Smith – arXiv preprint arXiv …, 2016 – arxiv.org
Page 1. Recurrent Neural Network Grammars Chris Dyer? Adhiguna Kuncoro? Miguel Ballesteros?? Noah A. Smith? … Abstract We introduce recurrent neural network gram- mars, probabilistic models of sentences with explicit phrase structure …

Deep reinforcement learning for dialogue generation
J Li, W Monroe, A Ritter, M Galley, J Gao… – arXiv preprint arXiv …, 2016 – arxiv.org
… To achieve these goals, we draw on the insights of reinforcement learning, which have been widely ap- plied in MDP and POMDP dialogue systems (see Re- lated Work section for details) … 2 Related Work Efforts to build statistical dialog systems fall into two major categories …

End-to-end LSTM-based dialog control optimized with supervised and reinforcement learning
JD Williams, G Zweig – arXiv preprint arXiv:1606.01269, 2016 – arxiv.org
… 7 Conclusion This paper has taken a first step toward end-to- end learning of task-oriented dialog systems. Our approach is based on a recurrent neural network which maps from raw dialog history to distribu- tions over actions …

Continuously learning neural dialogue management
PH Su, M Gasic, N Mrksic, L Rojas-Barahona… – arXiv preprint arXiv …, 2016 – arxiv.org
… based Dialog State Track- ing with Recurrent Neural Networks. In Proc of SIG- dial. [Jurc?cek et al.2011] Filip Jurc?cek, Blaise Thomson, and Steve Young. 2011. Natural actor and belief critic: Reinforcement algorithm for learning parameters of dialogue systems modelled as …

Learning to respond with deep neural networks for retrieval-based human-computer conversation system
R Yan, Y Song, H Wu – Proceedings of the 39th International ACM SIGIR …, 2016 – dl.acm.org
… We use a bi-directional recurrent neural network to propagate information across words; a convolutional neural network layer further captures patterns of adjacent words. Then a matching layer combines the information in each individual sentence …

Learning end-to-end goal-oriented dialog
A Bordes, J Weston – arXiv preprint arXiv:1605.07683, 2016 – arxiv.org
… Evaluating prerequisite qualities for learning end-to-end dialog systems. In Proc … Word-based dialog state tracking with recurrent neural networks. In Proceedings of the 15th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL), pages 292–299 …

Counter-fitting word vectors to linguistic constraints
N Mrkši?, DO Séaghdha, B Thomson, M Gaši?… – arXiv preprint arXiv …, 2016 – arxiv.org
… As far as we are aware, there is no previous work on exploiting antonymy in dialogue systems … In this paper we adopt the recurrent neural network (RNN) framework for tracking suggested in (Hender- son et al., 2014d; Henderson et al., 2014c; Mrkšic et al., 2015) …

Attention-based recurrent neural network models for joint intent detection and slot filling
B Liu, I Lane – arXiv preprint arXiv:1609.01454, 2016 – arxiv.org
… based neural network model for joint intent detection and slot filling, both of which are critical steps for many speech understanding and dialog systems … Index Terms: Spoken Language Understanding, Slot Filling, Intent Detection, Recurrent Neural Networks, Attention Model …

Dialog-based language learning
JE Weston – Advances in Neural Information Processing Systems, 2016 – papers.nips.cc
Paper accepted and presented at the Neural Information Processing Systems Conference (http://nips.cc/).

Towards end-to-end learning for dialog state tracking and management using deep reinforcement learning
T Zhao, M Eskenazi – arXiv preprint arXiv:1606.02560, 2016 – arxiv.org
… Their approach treated a dialog system as a mapping problem between the dialog history and the system response … The recurrent neural network can thus be viewed as an approximation of the belief state that can aggregate information from a sequence of observations …

On-line active reward learning for policy optimisation in spoken dialogue systems
PH Su, M Gasic, N Mrksic, L Rojas-Barahona… – arXiv preprint arXiv …, 2016 – arxiv.org
… a recurrent neural network encoder-decoder. The experimental results demonstrate that the proposed framework is able to significantly reduce data annota- tion costs and mitigate noisy user feedback in dialogue policy learning. 1 Introduction Spoken Dialogue Systems (SDS …

Contextual lstm (clstm) models for large scale nlp tasks
S Ghosh, O Vinyals, B Strope, S Roy, T Dean… – arXiv preprint arXiv …, 2016 – arxiv.org
… In this paper, we present CLSTM (Con- textual LSTM), an extension of the recurrent neural network LSTM (Long-Short Term … a wide variety of NL applications like question answering, sentence completion, paraphrase generation, and next utterance prediction in dialog systems …

A comparative study of recurrent neural network models for lexical domain classification
S Ravuri, A Stolcke – Acoustics, Speech and Signal Processing …, 2016 – ieeexplore.ieee.org
… Utterance classification is an important pre-processing step for many dialog systems that interpret speech input … in which a system must identify whether speech is directed at the machine, or another hu- man, and recently, we compared recurrent neural network (RNN) and long …

Deep recurrent models with fast-forward connections for neural machine translation
J Zhou, Y Cao, X Wang, P Li, W Xu – arXiv preprint arXiv:1606.04199, 2016 – arxiv.org
… Moreover, NMT models can also be easily adapted to other tasks such as dialog systems (Vinyals and Le, 2015), question answering systems (Yu et al … The recurrent neural network (RNN), or its spe- cific form the LSTM, is generally used as the basic unit of the encoding and …

Conditional generation and snapshot learning in neural dialogue systems
TH Wen, M Gasic, N Mrksic… – arXiv preprint arXiv …, 2016 – arxiv.org
… ar- chitectures and different ways to represent and aggregate the source information in an end- to-end neural dialogue system framework … Recurrent Neural Network (RNN)-based condi- tional language models (LM) have been shown to be very effective in solving a number of …

Sequence-to-sequence learning as beam-search optimization
S Wiseman, AM Rush – arXiv preprint arXiv:1606.02960, 2016 – arxiv.org
… also proven to be useful for sen- tence compression (Filippova et al., 2015), parsing (Vinyals et al., 2015), and dialogue systems (Ser- ban … y1:T to refer to the gold (ie, correct) target word sequence for an input x. Most seq2seq systems utilize a recurrent neural network (RNN) for …

Key-value memory networks for directly reading documents
A Miller, A Fisch, J Dodge, AH Karimi, A Bordes… – arXiv preprint arXiv …, 2016 – arxiv.org
… et al., 2002) have been recently re- visited (Tsai et al., 2015), the best published results on TRECQA and WIKIQA have been obtained by either convolutional neural networks (Santos et al., 2016; Yin and Schütze, 2015; Wang et al., 2016) or recurrent neural networks (Miao et al …

The dialog state tracking challenge series: A review
J Williams, A Raux, M Henderson – Dialogue & Discourse, 2016 – dad.uni-bielefeld.de
… Early spoken dialog systems used hand-crafted rules for dialog state tracking … Third, recurrent neural networks can be estimated where the inputs are the observed ASR/SLU results, and the output is a distribution over dialog states (Henderson et al., 2014d). Henderson et al …

Learning distributed representations of sentences from unlabelled data
F Hill, K Cho, A Korhonen – arXiv preprint arXiv:1602.03483, 2016 – arxiv.org
… Examples include machine translation (Sutskever et al., 2014), image captioning (Mao et al., 2015) and dialogue systems (Serban et al., 2015) … As with all sequence-to-sequence models, in train- ing the source sentence is ‘encoded’ by a Recurrent Neural Network (RNN) (with …

Generative deep neural networks for dialogue: A short review
IV Serban, R Lowe, L Charlin, J Pineau – arXiv preprint arXiv:1611.06216, 2016 – arxiv.org
… Multiresolution recurrent neural networks: An application to dialogue response generation. arXiv preprint arXiv:1606.00776, 2016a. IV Serban, A. Sordoni, Y. Bengio, AC Courville, and J. Pineau. Building end-to-end dialogue systems using generative hierarchical neural …

Multi-Domain Joint Semantic Frame Parsing Using Bi-Directional RNN-LSTM.
D Hakkani-Tür, G Tür, A Celikyilmaz… – …, 2016 – pdfs.semanticscholar.org
… [4] Y.-N. Chen, WY Wang, and AI Rudnicky, “Unsupervised in- duction and filling of semantic slots for spoken dialogue systems using frame-semantic parsing,” in Proceedings of ASRU … 22, no. 4, April 2014. [19] S. Ravuri and A. Stolcke, “Recurrent neural network and LSTM …

A sequence-to-sequence model for user simulation in spoken dialogue systems
LE Asri, J He, K Suleman – arXiv preprint arXiv:1607.00070, 2016 – arxiv.org
… Abstract User simulation is essential for generating enough data to train a statistical spoken dialogue system. Previous … This paper introduces a data-driven user simulator based on an encoder-decoder recurrent neural network. The …

End-to-End Memory Networks with Knowledge Carryover for Multi-Turn Spoken Language Understanding.
YN Chen, D Hakkani-Tür, G Tür, J Gao… – …, 2016 – pdfs.semanticscholar.org
… Spoken language understanding (SLU) is a core component of a spoken dialogue system … ex- tract salient semantics for modeling knowledge carryover in the multi-turn conversations and outperform the results using the state-of-the-art recurrent neural network framework (RNN …

Neural belief tracker: Data-driven dialogue state tracking
N Mrkši?, DO Séaghdha, TH Wen, B Thomson… – arXiv preprint arXiv …, 2016 – arxiv.org
… Abstract Belief tracking is a core component of modern spoken dialogue system pipelines. However … 1 Introduction Spoken dialogue systems (SDS) allow users to inter- act with computer applications through conversation. Task …

Dialogue State Tracking using Long Short Term Memory Neural Networks
K Yoshino, T Hiraoka, G Neubig, S Nakamura – 2016 – colips.org
… change expands the variety of expressions of users, because the users will be free from limitations im- posed by dialogue systems … Previously, recurrent neural networks (RNNs) have been used for the dialogue state tracking [2, 1] and they achieved good results in previous …

Question detection from acoustic features using recurrent neural network with gated recurrent unit
Y Tang, Y Huang, Z Wu, H Meng… – Acoustics, Speech and …, 2016 – ieeexplore.ieee.org
… Second, in most spoken dialog systems, automatic speech recognition (ASR) is the foremost step whose performance will have huge … Recurrent neural networks (RNN) can model context information along time steps of sequence [9]. Inspired by this characteristic, we propose an …

A neural knowledge language model
S Ahn, H Choi, T Pärnamaa, Y Bengio – arXiv preprint arXiv:1608.00318, 2016 – arxiv.org
… Specifically, in the recurrent neural network language model (RNNLM) [26] the computational complexity is linearly dependent on the number of vocabulary … many language/speech tasks such as speech recognition [26, 1], machine translation [17], and dialogue systems [40, 34] …

Neural Utterance Ranking Model for Conversational Dialogue Systems.
M Inaba, K Takahashi – SIGDIAL Conference, 2016 – aclweb.org
… To tackle this prob- lem, some studies extract correct sentences as ut- terances for dialogue systems from web data (In- aba et … Our proposed method processes the word se- quences in utterances and utterance sequences in context via multiple recurrent neural networks (RNNs …

Cuni system for wmt16 automatic post-editing and multimodal translation tasks
J Libovický, J Helcl, M Tlustý, P Pecina… – arXiv preprint arXiv …, 2016 – arxiv.org
… 2015. Sequence level training with recurrent neural networks. CoRR, abs/1511.06732. [Rosa2014] Rudolf Rosa. 2014 … 2015. Semantically conditioned lstm-based natural language generation for spoken dialogue systems …

Sequence to backward and forward sequences: A content-introducing approach to generative short-text conversation
L Mou, Y Song, R Yan, G Li, L Zhang, Z Jin – arXiv preprint arXiv …, 2016 – arxiv.org
… researchers have applied data-driven approaches, including retrieval methods (Isbell et al., 2000; Wang et al., 2013), phrase-based machine translation (Ritter et al., 2011), and recurrent neural networks (Sordoni et … Li et al. (2016b) propose a proactive dialogue system that can …

The dialogue breakdown detection challenge: Task description, datasets, and evaluation metrics.
R Higashinaka, K Funakoshi, Y Kobayashi, M Inaba – LREC, 2016 – lrec-conf.org
… Fatal or not? Finding errors that lead to dialogue breakdowns in chat-oriented dialogue systems. In Proc … (in Japanese). Kobayashi, S., Unno, Y., and Fukuda, M. (2015). Multi- task learning of recurrent neural network for detecting breakdowns of dialog and language modeling …

Listen, Attend, and Walk: Neural Mapping of Navigational Instructions to Action Sequences.
H Mei, M Bansal, MR Walter – AAAI, 2016 – aaai.org
… We propose a recurrent neural network with long short- term memory (LSTM) (Hochreiter and Schmidhuber 1997) to both encode the navigational instruction sequence bidi- rectionally and to decode the representation to an action sequence, based on a representation of the …

Implicit distortion and fertility models for attention-based encoder-decoder NMT model
S Feng, S Liu, M Li, M Zhou – arXiv preprint arXiv:1601.03317, 2016 – arxiv.org
Page 1. Implicit Distortion and Fertility Models for Attention-based Encoder-Decoder NMT Model Shi Feng† Shanghai Jiao Tong University Shanghai, PR China sjtufs@gmail.com Shujie Liu, Mu Li, Ming Zhou Microsoft Research …

Visual dialog
A Das, S Kottur, K Gupta, A Singh, D Yadav… – arXiv preprint arXiv …, 2016 – arxiv.org
… Hierarchical Recurrent Encoder: that contains a dialog- level Recurrent Neural Network (RNN) sitting on top of a question-answer (QA)-level recurrent block. In each … [13] is a statistical templated-question gener- ator and not an actual visual dialog system …

SampleRNN: An unconditional end-to-end neural audio generation model
S Mehri, K Kumar, I Gulrajani, R Kumar, S Jain… – arXiv preprint arXiv …, 2016 – arxiv.org
… The unreasonable effectiveness of recurrent neural networks. Andrej Karpathy blog, 2015 … Iulian V Serban, Alessandro Sordoni, Yoshua Bengio, Aaron Courville, and Joelle Pineau. Building end-to-end dialogue systems using generative hierarchical neural network models …

LSTM based Conversation Models
Y Luan, Y Ji, M Ostendorf – arXiv preprint arXiv:1603.09457, 2016 – arxiv.org
… Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems. arXiv preprint arXiv:1508.01745, 2015 … [13] Rui Lin, Shujie Liu, Muyun Yang, Mu Li, Ming Zhou, and Sheng Li. Hierarchical recurrent neural network for document modeling …

End-to-end reinforcement learning of dialogue agents for information access
B Dhingra, L Li, X Li, J Gao, YN Chen, F Ahmed… – arXiv preprint arXiv …, 2016 – arxiv.org
… All components of the KB- InfoBot are trained in an end-to-end fashion using reinforcement learning. Goal-oriented dialogue systems typically need to interact with an external database to access real-world knowledge (eg, movies playing in a city) …

Strategy and Policy Learning for Non-Task-Oriented Conversational Systems.
Z Yu, Z Xu, AW Black, AI Rudnicky – SIGDIAL Conference, 2016 – aclweb.org
… Ritter et al., 2011), retrieval-based response se- lection (Banchs and Li, 2012), and sequence-to- sequence recurrent neural network (Vinyals and … In a stochastic envi- ronment, a dialog system’s actions are system ut- terances, and the state is represented by the dialog history …

Summarizing Source Code using a Neural Attention Model.
S Iyer, I Konstas, A Cheung, L Zettlemoyer – ACL (1), 2016 – aclweb.org
… These approaches are not learning based, and require significant manual template-engineering efforts. We use recurrent neural networks (RNN) based on LSTMs and neural attention to jointly model source code and NL … (2015) generate text for spoken dialogue systems with a …

Adobe-MIT submission to the DSTC 4 Spoken Language Understanding pilot task
F Dernoncourt, JY Lee, TH Bui, HH Bui – arXiv preprint arXiv:1605.02129, 2016 – arxiv.org
… The Fourth Dialog State Tracking Challenge. In Proceedings of the 7th International Workshop on Spoken Dialogue Systems (IWSDS), 2016. 8. T. Kudo and Y. Matsumoto … Using recurrent neural networks for slot filling in spoken language understanding …

Improving neural language models with a continuous cache
E Grave, A Joulin, N Usunier – arXiv preprint arXiv:1612.04426, 2016 – arxiv.org
… Evaluating prerequisite qualities for learning end-to-end dialog systems. arXiv preprint arXiv:1511.06931, 2015. John Duchi, Elad Hazan, and Yoram Singer … A. Graves, A. Mohamed, and G. Hinton. Speech recognition with deep recurrent neural networks. In ICASSP, 2013 …

Joint Online Spoken Language Understanding and Language Modeling with Recurrent Neural Networks
B Liu, I Lane – arXiv preprint arXiv:1609.01462, 2016 – arxiv.org
Page 1. Joint Online Spoken Language Understanding and Language Modeling with Recurrent Neural Networks … Abstract Speaker intent detection and semantic slot filling are two critical tasks in spoken lan- guage understanding (SLU) for dialogue systems …

A Simple, Fast Diverse Decoding Algorithm for Neural Generation
J Li, W Monroe, D Jurafsky – arXiv preprint arXiv:1611.08562, 2016 – arxiv.org
Page 1. A Simple, Fast Diverse Decoding Algorithm for Neural Generation Jiwei Li, Will Monroe and Dan Jurafsky Computer Science Department, Stanford University, Stanford, CA, USA jiweil,wmonroe4,jurafsky@stanford.edu Abstract …

Polyglot neural language models: A case study in cross-lingual phonetic representation learning
Y Tsvetkov, S Sitaram, M Faruqui, G Lample… – arXiv preprint arXiv …, 2016 – arxiv.org
… To overcome the notorious problem in recurrent neural networks of vanishing gradients (Bengio et al., 1994), following Sundermeyer et al … Text-to-speech (TTS) sys- tems are also used as part of speech-to-speech trans- lation systems and spoken dialog systems, such as …

GuessWhat?! Visual object discovery through multi-modal dialogue
H de Vries, F Strub, S Chandar, O Pietquin… – arXiv preprint arXiv …, 2016 – arxiv.org
… Abstract We introduce GuessWhat?!, a two-player guessing game as a testbed for research on the interplay of computer vision and dialogue systems … Although goal-directed dialogue systems are appeal- ing, they remain hard to design …

Context-aware Natural Language Generation with Recurrent Neural Networks
J Tang, Y Yang, S Carton, M Zhang, Q Mei – arXiv preprint arXiv …, 2016 – arxiv.org
… The gradient is clipped if the norm is larger than 5. Dropout is used from the input to hidden layer and from the hidden layer to output layer in the recurrent neural networks … Oh, AH, and Rudnicky, AI 2000. Stochastic lan- guage generation for spoken dialogue systems …

Your paper has been accepted, rejected, or whatever: Automatic generation of scientific paper reviews
A Bartoli, A De Lorenzo, E Medvet, F Tarlao – International Conference on …, 2016 – Springer
… Lang. Eng. 14(4), 431–455 (2008)CrossRef. 22. Rieser, V., Lemon, O.: Natural language generation as planning under uncertainty for spoken dialogue systems … Mikolov, T., Karafiát, M., Burget, L., Cernock?, J., Khudanpur, S.: Recurrent neural network based language model …

Controlling output length in neural encoder-decoders
Y Kikuchi, G Neubig, R Sasano, H Takamura… – arXiv preprint arXiv …, 2016 – arxiv.org
Page 1. Controlling Output Length in Neural Encoder-Decoders Yuta Kikuchi1 kikuchi@lr.pi.titech. ac.jp Graham Neubig2? gneubig@cs.cmu.edu Ryohei Sasano1 sasano@pi.titech.ac.jp Hiroya Takamura1 takamura@pi.titech.ac.jp Manabu Okumura1 oku@pi.titecjh.ac.jp …

Dynamic Neural Turing Machine with Soft and Hard Addressing Schemes
C Gulcehre, S Chandar, K Cho, Y Bengio – arXiv preprint arXiv …, 2016 – arxiv.org
… The NTM consists of two main modules, a controller and, a memory. The controller, which is often implemented as a recurrent neural network, issues a command to the memory so as to read, write to and erase a subset of memory cells …

Sequence-based Structured Prediction for Semantic Parsing.
C Xiao, M Dymetman, C Gardent – ACL (1), 2016 – aclweb.org
… Given the recently shown effectiveness of RNNs (Recurrent Neural Networks), in particu- lar Long Short Term Memory (LSTM) networks (Hochreiter and Schmidhuber, 1997), for perform- ing sequence prediction in NLP applications such as machine translation (Sutskever et al …

A unified approach to transfer learning of deep neural networks with applications to speaker adaptation in automatic speech recognition
Z Huang, SM Siniscalchi, CH Lee – Neurocomputing, 2016 – Elsevier
In this paper, we present a unified approach to transfer learning of deep neural networks (DNNs) to address performance degradation issues caused by a potential.

Sequence-level knowledge distillation
Y Kim, AM Rush – arXiv preprint arXiv:1606.07947, 2016 – arxiv.org
… ap- proaches. NMT systems directly model the proba- bility of the next word in the target sentence sim- ply by conditioning a recurrent neural network on the source sentence and previously generated target words. While both …

NewsQA: A Machine Comprehension Dataset
A Trischler, T Wang, X Yuan, J Harris, A Sordoni… – arXiv preprint arXiv …, 2016 – arxiv.org
… How not to evaluate your dialogue system: An empirical study of unsupervised evaluation metrics for dialogue response generation. arXiv preprint arXiv:1603.08023, 2016 … On the difficulty of training recurrent neural networks. ICML (3), 28:1310–1318, 2013 …

Distributional semantics for understanding spoken meal descriptions
M Korpusik, C Huang, M Price… – Acoustics, Speech and …, 2016 – ieeexplore.ieee.org
… ASRU, Dec. 2011. [24] M. Korpusik, “Spoken language understanding in a nutrition dialogue system,” MS thesis, Massachusetts Institute of Tech- nology, 2015 … ACL, 2014. [30] G. Mesnil, X. He, L. Deng, and Y. Bengio, “Investigation of recurrent-neural-network architectures and …

Deep reinforcement learning for multi-domain dialogue systems
H Cuayáhuitl, S Yu, A Williamson, J Carse – arXiv preprint arXiv …, 2016 – arxiv.org
… include the following. [6] uses a Recurrent Neural Network (RNN) for dialogue act prediction in a POMDP-based dialogue system, which focuses on mapping system and user sentences to dialogue acts. [2] applies Deep Reinforcement …

Multi-modal variational encoder-decoders
IV Serban, II Ororbia, G Alexander, J Pineau… – arXiv preprint arXiv …, 2016 – arxiv.org
… Thus, for complex, multi-modal distributions — such as the distribution over topics in a text corpus, or natural language responses in a dialogue system — the uni-modal Gaussian prior inhibits the model’s ability to ex- tract and represent important structure in the data …

Knowledge as a teacher: Knowledge-guided structural attention networks
YN Chen, D Hakkani-Tur, G Tur, A Celikyilmaz… – arXiv preprint arXiv …, 2016 – arxiv.org
… Abstract Natural language understanding (NLU) is a core component of a spoken dialogue system. Recently recurrent neural networks (RNN) ob- tained strong results on NLU due to their supe- rior ability of preserving sequential informa- tion over time …

Deep Learning of Audio and Language Features for Humor Prediction.
D Bertero, P Fung – LREC, 2016 – lrec-conf.org
… 2.4. Recurrent Neural Network (RNN) The RNN is a neural network layout that provides a mem- ory component to the classifier, in the form of a … Our ultimate goal is to integrate laughter response prediction in a machine dialog system, to allow it to understand and react to humor …

Semantic language models with deep neural networks
AO Bayer, G Riccardi – Computer Speech & Language, 2016 – Elsevier
… Although currently the state-of-the-art performance for LMs is obtained by using recurrent neural networks (RNNs), n-gram LMs are still widely used … This may lead to problems especially for spoken dialog systems, where one of the main goals of these systems is to extract user …

A regression approach to single-channel speech separation via high-resolution deep neural networks
J Du, Y Tu, LR Dai, CH Lee – IEEE/ACM Transactions on Audio, Speech …, 2016 – dl.acm.org
… One category discusses the segregation of the background noises from the speech, including the noise perturbation for DNN-based speech separation [37], the use of long short-term memory based recurrent neural networks (RNNs) [38] and deep NMF models [39] …

Dialogue learning with human-in-the-loop
J Li, AH Miller, S Chopra, MA Ranzato… – arXiv preprint arXiv …, 2016 – arxiv.org
Page 1. Under review as a conference paper at ICLR 2017 DIALOGUE LEARNING WITH HUMAN-IN-THE-LOOP Jiwei Li, Alexander H. Miller, Sumit Chopra, Marc’Aurelio Ranzato, Jason Weston Facebook AI Research, New …

Chinese poetry generation with planning based neural network
Z Wang, W He, H Wu, H Wu, W Li, H Wang… – arXiv preprint arXiv …, 2016 – arxiv.org
… 2016. How not to evaluate your dialogue system: An empirical study of unsupervised evaluation metrics for dialogue response generation. arXiv preprint arXiv:1603.08023 … 2010. Recurrent neural network based language model. In INTERSPEECH, volume 2, page 3 …

Translating player dialogue into meaning representations using LSTMs
J Ryan, AJ Summerville, M Mateas… – … Conference on Intelligent …, 2016 – Springer
… to generate dialogue for our target application, a videogame, and train a long short-term memory (LSTM) recurrent neural network (RNN) to … In service dialogue systems, interaction is constrained and highly structured, lending well to rule-based approaches to natural language …

Log-linear rnns: Towards recurrent neural networks with flexible prior knowledge
M Dymetman, C Xiao – arXiv preprint arXiv:1607.02467, 2016 – arxiv.org
… 1 Introduction Recurrent Neural Networks (Goodfellow et al., 2016, Chapter 10) have re- cently shown remarkable success in sequential data prediction and have been applied to such NLP tasks as Language Modelling (Mikolov et al., 2010), 1 …

End-to-end joint learning of natural language understanding and dialogue manager
X Yang, YN Chen, D Hakkani-Tur, P Crook, X Li… – arXiv preprint arXiv …, 2016 – arxiv.org
… recurrent neural networks,” in SIGDIAL, 2014, pp. 292–299. [23] M Gašic, Catherine Breslin, Matthew Henderson, Dongho Kim, Martin Szummer, Blaise Thomson, Pirros Tsiakoulis, and Steve Young, “On-line policy optimi- sation of Bayesian spoken dialogue systems via human …

Dialog state tracking, a machine reading approach using Memory Network
J Perez, F Liu – arXiv preprint arXiv:1606.04052, 2016 – arxiv.org
… In the DSTC-2 dialog corpus, a user queries a database of local restaurants by interacting with a dialog system … We compare with two utterance-level discriminative neural trackers, a Recurrent Neural Network (RNN) model [8] and the Neural Belief Tracker [18] …

Multi-view Response Selection for Human-Computer Conversation.
X Zhou, D Dong, H Wu, S Zhao, D Yu, H Tian, X Liu… – EMNLP, 2016 – aclweb.org
… In the view of utter- ance sequence, discourse information can be learnt through utterance-level recurrent neural network, different from word sequence view … 2014. Empirical evaluation of gated recurrent neural networks on sequence model- ing. CoRR, abs/1412.3555 …

Incorporating Loose-Structured Knowledge into LSTM with Recall Gate for Conversation Modeling
Z Xu, B Liu, B Wang, C Sun, X Wang – arXiv preprint arXiv:1605.05110, 2016 – arxiv.org
… Indeed, the Recurrent Neural Network (RNN) and Long Short-term Mem- ory (LSTM) (Hochreiter and Schmidhuber, 1997) based methods have … context-aware queries and further select best answers based on the conversation his- tory, for building automatic dialog systems …

Syntax or semantics? knowledge-guided joint semantic frame parsing
YN Chen, D Hakanni-Tür, G Tur… – … (SLT), 2016 IEEE, 2016 – ieeexplore.ieee.org
… Spoken language understanding (SLU) is a core component of a spo- ken dialogue system, which involves intent prediction and slot filling and also called semantic frame parsing. Recently recurrent neural networks (RNN) obtained strong results on SLU due to their superior …

Recurrent Neural Networks for Dialogue State Tracking
O Plátek, P B?lohlávek, V Hude?ek… – arXiv preprint arXiv …, 2016 – arxiv.org
… and compared two dialogue state tracking models which are based on state-of-the-art architectures using recurrent neural networks … Informal experiments were conducted during the Statistical Dialogue Systems course at Charles University (see https://github.com/oplatek/sds …

Online Sequence-to-Sequence Reinforcement Learning for Open-Domain Conversational Agents
N Asghar, P Poupart, J Xin, H Li – arXiv preprint arXiv:1612.03929, 2016 – arxiv.org
… Building end-to-end dialogue systems using generative hi- erarchical neural network models. In AAAI, pages 3776– 3784, 2016 … Multiresolution recurrent neural networks: An application to dialogue response gen- eration. arXiv preprint arXiv:1606.00776, 2016 …

Dialogue session segmentation by embedding-enhanced texttiling
Y Song, L Mou, R Yan, L Yi, Z Zhu, X Hu… – arXiv preprint arXiv …, 2016 – arxiv.org
… 2.1. Dialogue Systems and Context Modeling Human-computer dialogue systems can be roughly divided into several categories … Sordoni et al. [12] summarize a single previous sentence as bag-of-words features, which are fed to a recurrent neural network for reply generation …

Natural language generation in dialogue using lexicalized and delexicalized data
S Sharma, J He, K Suleman, H Schulz… – arXiv preprint arXiv …, 2016 – arxiv.org
… Abstract Natural language generation plays a critical role in any spoken dialogue system. We present a new approach to natural language generation using recurrent neural networks in an encoder- decoder framework. In contrast …

Knowledge-based semantic embedding for machine translation
C Shi, S Liu, S Ren, S Feng, M Li, M Zhou… – Proceedings of the …, 2016 – aclanthology.info
… method. In our KB- SE method, the source sentence is firstly mapped into a knowledge based seman- tic space, and the target sentence is gen- erated using a recurrent neural network with the internal meaning preserved. Ex …

Dialogue Act Classification in Domain-Independent Conversations Using a Deep Recurrent Neural Network.
H Khanpour, N Guntakandla, R Nielsen – COLING, 2016 – aclweb.org
… Many applications benefit from the use of automatic dialogue act classi- fication such as dialogue systems, machine translation, Automatic … Lee and Dernoncourt, 2016; Kalchbrenner and Blunsom, 2013), we propose a model based on a recurrent neural network, LSTM, that …

Exploring Convolutional and Recurrent Neural Networks in Sequential Labelling for Dialogue Topic Tracking.
S Kim, RE Banchs, H Li – ACL (1), 2016 – aclweb.org
… topic tracking, including convolutional neural networks to account for semantics at each individual utterance, and recurrent neural networks to account for … These human capabilities for handling topics are also expected from dialogue systems to achieve natural and human-like …

Conversational Contextual Cues: The Case of Personalization and History for Response Ranking
R Al-Rfou, M Pickett, J Snaider, Y Sung… – arXiv preprint arXiv …, 2016 – arxiv.org
… such as games (Narasimhan et al., 2015) and restaurants (Wen et al., 2016; Cuayáhuitl, 2016) Personalizing dialogue systems requires sufficient … to- sequence framework (Sutskever et al., 2014), many recent learning systems have used recurrent neural networks (RNNs) to …

An overview of end-to-end language understanding and dialog management for personal digital assistants
R Sarikaya, PA Crook, A Marin, M Jeong… – … (SLT), 2016 IEEE, 2016 – ieeexplore.ieee.org
… Khan, JP Robichaud, P. Crook, R. Sarikaya, Hypotheses Ranking and State Tracking for a Multi-Domain Dialog System using ASR … 15] P. Xu and R. Sarikaya, “Contextual domain classification in spoken language understanding systems using recurrent neural network, in Proc …

An attentional neural conversation model with improved specificity
K Yao, B Peng, G Zweig, KF Wong – arXiv preprint arXiv:1606.01292, 2016 – arxiv.org
Page 1. An Attentional Neural Conversation Model with Improved Specificity Kaisheng Yao Microsoft Research Redmond, USA kaisheny@microsoft.com Baolin Peng Chinese University of Hong Kong blpeng@se.cuhk.edu.hk …

A Sentence Interaction Network for Modeling Dependence between Sentences.
B Liu, M Huang, S Liu, X Zhu, X Zhu – ACL (1), 2016 – aclweb.org
… While deep learning methods like Recurrent Neural Network or Convo- lutional Neural Network have been proved to be powerful for sentence modeling, prior studies paid less attention on inter- actions between sentences …

Dialog History Construction with Long-Short Term Memory for Robust Generative Dialog State Tracking
BJ Lee, KE Kim – Dialogue & Discourse, 2016 – dad.uni-bielefeld.de
… is heavily affected by that of the dialog tracker, it has been one of the core areas of research on dialog systems. In this paper, we present a dialog state tracker that combines a generative probabilistic model of dialog state tracking with the recurrent neural network for encoding …

Task Lineages: Dialog State Tracking for Flexible Interaction.
S Lee, A Stent – SIGDIAL Conference, 2016 – aclweb.org
… Session-based, single task, simple goal dialog is eas- ier for dialog system engineers and consistent with 25 years of commercial dialog system development, but does not match users’ real-world task needs as communicated with human conversational as- sistants or …

Context-Sensitive and Role-Dependent Spoken Language Understanding Using Bidirectional and Attention LSTMs.
C Hori, T Hori, S Watanabe, JR Hershey – INTERSPEECH, 2016 – merl.com
… Spoken language under- standing (SLU) technologies in dialog systems have been inten- sively investigated to estimate the intention of user … Recurrent neural networks (RNNs) have been more actively applied to utterance classification to consider history of a word sequence in …

Dialog state tracking with attention-based sequence-to-sequence learning
T Hori, H Wang, C Hori, S Watanabe… – … (SLT), 2016 IEEE, 2016 – ieeexplore.ieee.org
… SLU) technology, which predicts the intention of spoken user utterances, is a key component of dialog systems [1, 2 … Support vec- tor machines (SVMs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs) have been applied to utterance- level tagging …

Overview of NTCIR-13
MP Kato, Y Liu, C Gurrin, H Joho… – Proceedings of the …, 2016 – research.nii.ac.jp
… We used recurrent neural networks (RNNs) to extract the answer … Each dataset has no less than 60 million pairs of text, and we aim to show how effective these combinations of large-scale datasets and large-scale neural models are for developing dialog systems …

Seeing is believing: the quest for multimodal knowledge by Gerard de Melo and Niket Tandon, with Martin Vesely as coordinator
G de Melo, N Tandon – ACM SIGWEB Newsletter, 2016 – dl.acm.org
… In speech recognition, markedly lower error rates have enabled powerful dialog systems, including Siri on Apple’s iOS, Alexa for Amazon Echo, and … While sufficient for traditional statistical NLP methods of the past, it turns out that deep recurrent neural networks need even more …

Chatbot evaluation and database expansion via crowdsourcing
Z Yu, Z Xu, AW Black… – Proceedings of the RE …, 2016 – pdfs.semanticscholar.org
… Unlike goal oriented dialog systems, chat- bots do not have any specific goal that guides the interac- tion … re- sponses, such as machine translation (Ritter et al., 2011), retrieval based response selection (Banchs and Li, 2012), and recurrent neural network sequence generation …

Assisting discussion forum users using deep recurrent neural networks
JHP Suorra, O Mogren – Proceedings of the 1st Workshop on …, 2016 – aclweb.org
… 2015. Docu- ment modeling with gated recurrent neural network for sentiment classification. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1422–1432 … 2013. Pomdp-based statistical spoken dialog systems: A review …

Leveraging Sentence-level Information with Encoder LSTM for Semantic Slot Filling
G Kurata, B Xiang, B Zhou, M Yu – arXiv preprint arXiv:1601.01530, 2016 – arxiv.org
… 2015. Deep contextual language under- standing in spoken dialogue systems. In Proc. INTER- SPEECH, pages 120–124 … 2015. Recurrent neural networks with external mem- ory for language understanding. arXiv preprint arXiv:1506.00195. [Price1990] Patti Price. 1990 …

A Wizard-of-Oz Study on A Non-Task-Oriented Dialog Systems That Reacts to User Engagement.
Z Yu, L Nicolich-Henkin, AW Black, AI Rudnicky – SIGDIAL Conference, 2016 – aclweb.org
… cO2016 Association for Computational Linguistics A Wizard-of-Oz Study on A Non-Task-Oriented Dialog Systems That Reacts to … Ritter et al., 2011), retrieval-based response selection (Banchs and Li, 2012), and sequence-to-sequence recurrent neural network (Vinyals and Le …

Encoder-decoder with Focus-mechanism for Sequence Labelling Based Spoken Language Understanding
S Zhu, K Yu – arXiv preprint arXiv:1608.02097, 2016 – arxiv.org
… In spoken dialogue system, the Spoken Language Under- standing (SLU) is a key component that parses user utter- ances into … learning approaches [5, 6], many neural network architectures have been applied to this task, such as simple recurrent neural networks (RNNs) [7, 8 …

CFGs-2-NLU: Sequence-to-sequence learning for mapping utterances to semantics and pragmatics
AJ Summerville, J Ryan, M Mateas… – arXiv preprint arXiv …, 2016 – arxiv.org
… to generate dialogue for our target application for NLU, a videogame, and train a long short-term memory (LSTM) recurrent neural network (RNN) to … But while service dialogue systems have become common, general conversational agents are still an open area of research …

A Joint Model of Intent Determination and Slot Filling for Spoken Language Understanding.
X Zhang, H Wang – IJCAI, 2016 – pdfs.semanticscholar.org
… Recurrent neural networks (RNNs) have been proved effective in SF, while there is no prior work using RNNs in ID … Spoken language understanding (SLU) in human/machine spoken dialog systems aims to automatically identify the in- tent of the user as expressed in natural …

Recurrent convolutional neural networks for structured speech act tagging
T Ushio, H Shi, M Endo, K Yamagami… – … Workshop (SLT), 2016 …, 2016 – ieeexplore.ieee.org
… paragraphs or longer texts using bag-of-n-grams or sentence vector averaging strategies [7 , 8]. KaIchbrenner et al.[9] have, however, proposed a sentence model based on convolutional neural networks(CNN) and a discourse model based on recurrent neural networks that is …

Dialog state tracking, a machine reading approach using a memory-enhanced neural network
J Perez – CoRR, abs/1606.04052, 2016 – pdfs.semanticscholar.org
… Word-based dialog state tracking with recurrent neural networks. In Proceedings of SIGDial, 2014. [JM15] Armand Joulin and Tomas Mikolov … Discriminative state tracking for spoken dialog systems. In Association for Computer Linguistics, pages 466–475 …

Neural Emoji Recommendation in Dialogue Systems
R Xie, Z Liu, R Yan, M Sun – arXiv preprint arXiv:1612.04609, 2016 – arxiv.org
… neural tensor network [Socher et al., 2013] to sentiment analysis, while [Tang et al., 2015] utilizes gated recurrent neural network to learn … our model is the first attempt on emoji classification by taking the contextual information into con- sideration in multi-turn dialogue systems …

Driver confusion status detection using recurrent neural networks
C Hori, S Watanabe, T Hori… – Multimedia and Expo …, 2016 – ieeexplore.ieee.org
… Fig. 1. Recurrent Neural Network. Fig. 2. LSTM cell. The input vector is projected to the D dimensional vector … v. 3, 2000, p. 362-365, 2000. [2] Teruhisa Misu, Antoine Raux, Ian Lane, Joan Devassy, and Rakesh Gupta, “Situated multi-modal dialog system in vehicles,” in …

Optimizing neural network hyperparameters with gaussian processes for dialog act classification
F Dernoncourt, JY Lee – Spoken Language Technology …, 2016 – ieeexplore.ieee.org
Page 1. OPTIMIZING NEURAL NETWORK HYPERPARAMETERS WITH GAUSSIAN PROCESSES FOR DIALOG ACT CLASSIFICATION Franck Dernoncourt*, Ji Young Lee * MITCSAIL Cambridge, MA, USA {francky,jjylee}@mit.edu ABSTRACT …

Two are Better than One: An Ensemble of Retrieval-and Generation-Based Dialog Systems
Y Song, R Yan, X Li, D Zhao, M Zhang – arXiv preprint arXiv:1610.07149, 2016 – arxiv.org
… Generative dialog systems, on the other hand, can synthesize a new sentence as the reply by language mod- els [16, 18, 19]. Typically, a recurrent neural network (RNN) captures the query’s semantics with one or a few distributed, real-valued vectors (also known as …

A hierarchical lstm model for joint tasks
Q Zhou, L Wen, X Wang, L Ma, Y Wang – China National Conference on …, 2016 – Springer
… For example, Chinese word segmentation and POS-tagging, POS-tagging and chunking, intent identification and slot filling in goal-driven spoken language dialogue systems, and so on … Shi et al. [8] proposed a hybrid model of Recurrent Neural Network (RNN) and …

Sample-efficient Deep Reinforcement Learning for Dialog Control
K Asadi, JD Williams – arXiv preprint arXiv:1612.06000, 2016 – arxiv.org
… Page 2. episodic problem, the return at a timestep t is: Gt = T?t ? i=1 ?i?1rt+i , (1) where T is the terminal timestep, and ? is a discount factor 0 ? ? ? 1. In this paper, we consider policies represented as a recurrent neural network (RNN) … 4 Problem 1: dialog system …

A step beyond local observations with a dialog aware bidirectional GRU network for Spoken Language Understanding
V Vukotic, C Raymond, G Gravier – Interspeech, 2016 – hal.inria.fr
… In the first part, we evaluate different recurrent neural network architectures, such as simple recurrent neural networks (RNN), long short-term … 2.2. MEDIA The research project MEDIA [13] evaluates different SLU mod- els of spoken dialogue systems dedicated to provide tourist in …

A context-aware natural language generator for dialogue systems
O Dušek, F Jur?í?ek – arXiv preprint arXiv:1608.07076, 2016 – arxiv.org
… neural networks with convolutional sentence reranking. In Proc. of SIGDIAL, pages 275–284. T.-H. Wen, M. Gasic, N. Mrkšic, P.-H. Su, D. Vandyke, and S. Young. 2015b. Semantically conditioned LSTM-based natural language generation for spo- ken dialogue systems. In Proc …

Character-Level Neural Translation for Multilingual Media Monitoring in the SUMMA Project
G Barzdins, S Renals, D Gosko – arXiv preprint arXiv:1604.01221, 2016 – arxiv.org
… stream segmentation into the stories it is possible to utilize the exceptional generalization and memorization capacity of the neural networks, which is already applied in the neural dialogue systems such as … Towards end-to-end speech recognition with recurrent neural networks …

A user simulator for task-completion dialogues
X Li, ZC Lipton, B Dhingra, L Li, J Gao… – arXiv preprint arXiv …, 2016 – arxiv.org
… We consider a dialogue system for helping users to book movie tickets or to look up the movies they want, by interacting with them in natural … The natural language understanding (NLU) component is a recurrent neural network model with long-short term memory (LSTM) cells …

Context-aware Natural Language Generation for Spoken Dialogue Systems.
H Zhou, M Huang, X Zhu – COLING, 2016 – aclweb.org
… 2013. Speech recognition with deep recurrent neural networks. In 2013 IEEE international conference on acoustics, speech and signal processing, pages 6645–6649 … Alice H Oh and Alexander I Rudnicky. 2000. Stochastic language generation for spoken dialogue systems …

Conversational In-Vehicle Dialog Systems: The past, present, and future
F Weng, P Angkititrakul, EE Shriberg… – IEEE Signal …, 2016 – ieeexplore.ieee.org
… putational power available through the cloud, more applica- tions and AI technologies started to be integrated into dialog systems; see Figure 2 … as model noise, reverberation, or speaker variation, in an end-to-end speech recognition training with a recurrent neural network [23] …

Reference-Aware Language Models
Z Yang, P Blunsom, C Dyer, W Ling – arXiv preprint arXiv:1611.01628, 2016 – arxiv.org
… How should the RE be rendered?) augment a traditional recurrent neural network language model and the two components are combined as a mixture … We first apply our model on task-oriented dialogue systems in the domain of restaurant recommenda- tions, and work on the …

Information density and overlap in spoken dialogue
N Dethlefs, H Hastie, H Cuayáhuitl, Y Yu… – Computer Speech & …, 2016 – Elsevier
… Our results are relevant for spoken dialogue systems, especially incremental ones. Abstract … 1. Introduction. Traditionally, the smallest unit of processing in spoken dialogue systems has been a full utterance with strict, rigid turn-taking …

Domain Adaptation for Neural Networks by Parameter Augmentation
Y Watanabe, K Hashimoto, Y Tsuruoka – arXiv preprint arXiv:1607.00410, 2016 – arxiv.org
… Recently, Recurrent Neural Networks (RNNs) have been successfully applied to various tasks in the field of natural language processing (NLP), in- cluding … (2016) have proposed a procedure to generate natural language for multiple domains of spoken dialogue systems …

LSTM Deep Neural Networks Postfiltering for Improving the Quality of Synthetic Voices
M Coto-Jiménez, J Goddard-Close – Mexican Conference on Pattern …, 2016 – Springer
… which aid the visually impaired, to in-car navigation systems, e-book readers, spoken dialog systems, communicative robots … 18–21], for example restricted boltzmann machines, deep belief networks, bidirectional associative memories, and recurrent neural networks (RNN) …

Large-scale acquisition of commonsense knowledge via a quiz game on a dialogue system
N Otani, D Kawahara, S Kurohashi, N Kaji… – OKBQA …, 2016 – pdfs.semanticscholar.org
… 2016. Design of word association games using dialog systems for acquisition of word association knowledge. In … 2015. Morphological analysis for unsegmented lan- guages using recurrent neural network language model. In …

JUNITMZ at SemEval-2016 Task 1: Identifying Semantic Similarity Using Levenshtein Ratio.
S Sarkar, D Das, P Pakray, AF Gelbukh – SemEval@ NAACL-HLT, 2016 – aclweb.org
… sim- ilarity are useful to a broad range of applica- tions including: text mining, information re- trieval, dialogue systems, machine translation and … 4http://nl.mathworks.com/help/nnet/ref/trainrp. html 5http://nl.mathworks.com/help/nnet/ug/design-layer- recurrent-neural-networks.html …

Length bias in encoder decoder models and a case for global conditioning
P Sountsov, S Sarawagi – arXiv preprint arXiv:1606.03402, 2016 – arxiv.org
Page 1. Length bias in Encoder Decoder Models and a Case for Global Conditioning Pavel Sountsov Google siege@google.com Sunita Sarawagi ? IIT Bombay sunita@iitb.ac.in Abstract Encoder-decoder networks are popular …

Exploiting Sentence and Context Representations in Deep Neural Models for Spoken Language Understanding
LMR Barahona, M Gasic, N Mrkši?, PH Su… – arXiv preprint arXiv …, 2016 – arxiv.org
… 2012. Statistical methods for building robust spoken dialogue systems in an automobile. Proceedings of the 4th applied human factors and ergonomics … 2013. Recurrent neural networks for language understanding. In INTERSPEECH, pages 2524–2528 …

A novel density-based clustering method using word embedding features for dialogue intention recognition
J Jang, Y Lee, S Lee, D Shin, D Kim, H Rim – Cluster Computing, 2016 – Springer
… 482–491 (2012). 17. Kang, S., Park, H., Seo, J.: Emotion classification of user’s utterance for a dialogue system. Korean J. Cognit … 3, 993–1022 (2003)MATHGoogle Scholar. 22. Mikolov, T., Karafiat, M., Burget, L., Cernocky, J.: Recurrent neural network based language model …

Improving generalisation to new speakers in spoken dialogue state tracking
I Casanueva, T Hain, P Green – Proceedings of the Annual …, 2016 – eprints.whiterose.ac.uk
… [16] B. Thomson, and S. Young. “Bayesian update of dialogue state: A POMDP framework for spoken dialogue systems”. Computer Speech and Language, 2010 … “Robust dialog state tracking using delexicalised recurrent neural networks and un- supervised adaptation” …

Contextual LSTM: A Step towards Hierarchical Language Modeling
S Ghosh, O Vinyals, B Strope, S Roy, T Dean, L Heck – 2016 – research.google.com
… In this paper, we present CLSTM (Contextual LSTM), an extension of the recurrent neural network LSTM (Long-Short Term Memory … variety of NL applications like question answering, sentence completion, paraphrase generation, and next utterance prediction in dialog systems …

Recent Advances on Human-Computer Dialogue
X Wang, C Yuan – CAAI Transactions on Intelligence Technology, 2016 – Elsevier
… For a more comprehensive survey on traditionally dialogue systems, especially on POMDP based pipeline dialogue systems, please read the excellent reviews by Young, Gasic … Conditional Random Field (CRF) and Recurrent Neural Network (RNN) were the mostly used models …

Medical examination data prediction using simple recurrent network and long short-term memory
HG Kim, GJ Jang, HJ Choi, M Kim, YW Kim… – Proceedings of the Sixth …, 2016 – dl.acm.org
… ABSTRACT In this work, we use two different types of recurrent neural networks (RNNs) to predict medical examination results of a subject given the previous measurements … Keywords medical examination data prediction; long short-term mem- ory; recurrent neural network …

Learning to Start for Sequence to Sequence Architecture
Q Zhu, W Zhang, L Zhou, T Liu – arXiv preprint arXiv:1608.05554, 2016 – arxiv.org
… 2013. On-line policy optimisation of bayesian spoken dialogue systems via human interaction … [Graves2012a] Alex Graves. 2012a. Neural networks. In Supervised Sequence Labelling with Recurrent Neural Networks, pages 15–35. Springer. [Graves2012b] Alex Graves. 2012b …

Predicting Dialogue Acts for Intelligent Virtual Agents with Multimodal Student Interaction Data.
W Min, JB Wiggins, L Pezzullo, AK Vail, KE Boyer… – EDM, 2016 – cise.ufl.edu
… The selection of dialogue moves was informed by the literature on dialogue systems for learning [8], as well as experience with a recent … 5.1 LSTM Background LSTMs are a type of gated recurrent neural network specifically designed for sequence labeling on temporal data …

Natural Language Generation through Character-based RNNs with Finite-state Prior Knowledge.
R Goyal, M Dymetman, E Gaussier, U LIG – COLING, 2016 – aclweb.org
… In particular, architectures based on Recurrent Neural Network (RNN) such as LSTMs (Hochreiter and Schmidhuber, 1997) and GRUs (Cho et al., 2014) have been succesfully used in Language Modelling tasks due to their … (2015) in the context of a dialog system, where the …

Learning to Query, Reason, and Answer Questions On Ambiguous Texts
X Guo, T Klinger, C Rosenbaum, JP Bigus, M Campbell… – 2016 – openreview.net
… There has been a lot of recent interest on the end-to-end training of dialog systems that are capable of generating a … For large corpora it is natural to use supervised training techniques where the Recurrent Neural Networks (RNNs) attempt to replicate the recorded human …

Statistical Natural Language Generation from Tabular Non-textual Data.
J Mahapatra, SK Naskar, S Bandyopadhyay – INLG, 2016 – anthology.aclweb.org
… An idea of generating text through recurrent neural network based approach with Hessian-free optimization was proposed by (Sutskever et al., 2011) … 2009. Mountain: a translation-based approach to natural language gener- ation for dialog systems …

Dynamic Neural Turing Machine with Continuous and Discrete Addressing Schemes
C Gulcehre, S Chandar, K Cho, Y Bengio – 2016 – openreview.net
… The NTM consists of two main modules, a controller and, a memory. The controller, which is often implemented as a recurrent neural network, issues a command to the memory so as to read, write to and erase a subset of memory cells …

RACAI Entry for the IWSLT 2016 Shared Task
S Pipa, AF Vasile, I Ionascu… – Proceedings of the …, 2016 – workshop2016.iwslt.org
… Page 2. (part-of-speech tagged, chunked and parsed). Text normalization is extremely important for automatic machine translation (MT), speech-to-speech translation, information extraction, dialog systems, etc … Word embedding for recurrent neural network based tts synthesis …

Spectral decomposition method of dialog state tracking via collective matrix factorization
J Perez – arXiv preprint arXiv:1606.05286, 2016 – arxiv.org
… Word-based dialog state tracking with recurrent neural networks. In in Proceedings of SIGdial, 2014c … Unsu- pervised spoken language understanding for a multi-domain dialog system. IEEE Transactions on Audio, Speech & Language Processing, 21(11):2451–2464, 2013 …

Towards Empathetic Human-Robot Interactions
P Fung, D Bertero, Y Wan, A Dey, RHY Chan… – arXiv preprint arXiv …, 2016 – arxiv.org
… only to find human users disappointed by the lack of reciprocal empathy from these robots. It follows that we shall embody interactive dialog systems in simulated or robotic forms … It is an improvement over the Recurrent Neural Network aimed to enhance its memory capabilities …

Controlling the voice of a sentence in japanese-to-english neural machine translation
H Yamagishi, S Kanouchi, T Sato… – Proceedings of the 3rd …, 2016 – anthology.aclweb.org
… 5 Related Work An NMT framework consists of two recurrent neural networks (RNNs), called the RNN encoder-decoder, proposed by Cho et al.(2014) and Sutskever et al.(2014) … For example, one may prefer a polite expression for generating conversation in a dialog system …

DeepSoft: A vision for a deep model of software
HK Dam, T Tran, J Grundy, A Ghose – Proceedings of the 2016 24th ACM …, 2016 – dl.acm.org
… DeepSoft leverages a deep recurrent neural network (RNN) to model this temporal evo- lution … Given the recent successes in NLP [5] (machine translation, question answering, and dialog systems) and vi- sion [4] (image/video captioning/story telling and more re- cently, visual …

Improving Attention Modeling with Implicit Distortion and Fertility for Machine Translation.
S Feng, S Liu, N Yang, M Li, M Zhou, KQ Zhu – COLING, 2016 – aclweb.org
… Alex Graves. 2013. Generating sequences with recurrent neural networks. arXiv preprint arXiv:1308.0850 … 2015. Se- mantically conditioned lstm-based natural language generation for spoken dialogue systems. arXiv preprint arXiv:1508.01745. Huijuan Xu and Kate Saenko …

Generating Paraphrases from DBPedia using Deep Learning
A Sleimi, C Gardent – WebNLG 2016, 2016 – webnlg2016.sciencesconf.org
… 2011. Generating text with recurrent neural networks. In Proceedings of the 28th International Conference on Machine Learning (ICML-11), pages 1017–1024 … 2015. Semantically conditioned ltsm-base natural lan- guage generation for spoken dialogue systems …

Towards Using Conversations with Spoken Dialogue Systems in the Automated Assessment of Non-Native Speakers of English.
DJ Litman, SJ Young, MJF Gales, K Knill… – SIGDIAL …, 2016 – aclweb.org
… International Workshop on Spoken Dialogue Systems. Nikola Mrkšic, Diarmuid O Séaghdha, Blaise Thom- son, Milica Gasic, Pei-Hao Su, David Vandyke, Tsung-Hsien Wen, and Steve Young. 2015. Multi- domain dialog state tracking using recurrent neural networks …

Coupling distributed and symbolic execution for natural language queries
L Mou, Z Lu, H Li, Z Jin – arXiv preprint arXiv:1612.02741, 2016 – arxiv.org
… MarcAurelio Ranzato, Sumit Chopra, Michael Auli, and Woj- ciech Zaremba. Sequence level training with recurrent neural networks. In ICLR, 2016 … A network-based end-to-end trainable task- oriented dialogue system. arXiv preprint arXiv:1604.04562, 2016 …

RNN-based Encoder-decoder Approach with Word Frequency Estimation
J Suzuki, M Nagata – arXiv preprint arXiv:1701.00138, 2016 – arxiv.org
… [2014], Cho et al. [2014] question answering Xu et al. [2015], dialogue system Vinyals and Le [2015], Shang et al … Sumit Chopra, Michael Auli, and Alexander M. Rush. Abstractive Sentence Summarization with Attentive Recurrent Neural Networks …

Learning interactive behavior for service robots the challenge of mixed-initiative interaction
P Liu, DF Glas, T Kanda, H Ishiguro – Proceedings of the Workshop on …, 2016 – irc.atr.jp
… and gaze behaviors were recognized in an imitative game using a hidden Markov model [8]. Data-driven dialogue systems have been … Recurrent neural networks are often used for tasks like language processing, and Long Short-Term Memory (LSTM) recurrent neural network …

Compressing neural language models by sparse word representations
Y Chen, L Mou, Y Xu, G Li, Z Jin – arXiv preprint arXiv:1610.03950, 2016 – arxiv.org
… from a neural LM, benefit- ing various downstream tasks like machine trans- lation, summarization, and dialogue systems (De- vlin … Recurrent neural networks (RNNs) can also be used for language modeling; they are especially capable of capturing long range dependencies in …

Learning Through Dialogue Interactions
J Li, AH Miller, S Chopra, MA Ranzato… – arXiv preprint arXiv …, 2016 – arxiv.org
… In the context of dialogue, with the recent popularity of deep learning models, many neural dialogue systems have been proposed … As far as we know, current dialogue systems mostly focus on learning through fixed supervised signals rather than interacting with users …

Exploiting turn-taking temporal evolution for personality trait perception in dyadic conversations
MH Su, CH Wu, YT Zheng – IEEE/ACM Transactions on Audio …, 2016 – ieeexplore.ieee.org
… The first phase includes the recurrent neural networks (RNNs) [23]–[25] and it entails generating the BFI item scores for each turn of a speaker (speaker turn). During a conver- sation, an individual spoken text might be projected onto a 10-dimensional point in the BFI space (Fig …

Manipulating Word Lattices to Incorporate Human Corrections.
Y Gaur, F Metze, JP Bigham – INTERSPEECH, 2016 – cs.cmu.edu
… [7] used it to transcribe podcasts, [8] used it to collect data from speech dialog systems and [9, 10] have used it to transcribe conversational speech … [19] W. Williams, N. Prasad, D. Mrva, T. Ash, and T. Robinson, “Scaling recurrent neural network language models,” in Acous- tics …

Sociocognitive Language Processing–Emphasising the Soft Factors
BW Schuller, MF McTear – … on Spoken Dialogue Systems (IWSDS …, 2016 – fim.uni-passau.de
… 1. Allen, J., Manshadi, M., Dzikovska, M., Swift, M.: Deep linguistic processing for spoken dialogue systems … Mesnil, G., Dauphin, Y., Yao, K., Bengio, Y., Deng, L., Hakkani-Tur, D., He, X., Heck, L., Tur, G., Yu, D., Zweig, G.: Using recurrent neural networks for slot filling in spoken …

JNDSLAM: A SLAM extension for speech synthesis
R Dall, X Gonzalvo – Proc. Speech Prosody, 2016 – researchgate.net
… This has potential uses in dialogue systems where we know what we wish to emphasise, whether we are asking a questions etc., phenomena which has specific … [28] H. Sak, AW Senior, and F. Beaufays, “Long short-term mem- ory recurrent neural network architectures for large …

Deep learning for sentiment analysis
LM Rojas?Barahona – Language and Linguistics Compass, 2016 – Wiley Online Library
… 3.3 Recurrent neural network. An NN that contains direct cycles in their hidden connections is a recurrent neural network (RNN). The hidden layer of an RNN is equal to … A simple recurrent neural network (RNN). Note that the recurrent connection connects two hidden layers …

Detecting paralinguistic events in audio stream using context in features and probabilistic decisions
R Gupta, K Audhkhasi, S Lee, S Narayanan – Computer Speech & …, 2016 – Elsevier
… interest. Potential techniques include Markov models (Rabiner and Juang, 1986), recurrent neural networks (Funahashi and Nakamura, 1993) and linear chain conditional random fields (Lafferty et al., 2001). For instance, Cai et al …

Symbol emergence in robotics: a survey
T Taniguchi, T Nagai, T Nakamura, N Iwahashi… – Advanced …, 2016 – Taylor & Francis
Page 1. ADVANCED ROBOTICS, 2016 VOL. 30, NOS. 11–12, 706–728 http://dx.doi.org/10.1080/01691864.2016.1164622 SURVEY PAPER Symbol emergence in robotics: a survey Tadahiro Taniguchia, Takayuki Nagaib, Tomoaki …

Leveraging Recurrent Neural Networks for Multimodal Recognition of Social Norm Violation in Dialog
T Zhao, R Zhao, Z Meng, J Cassell – arXiv preprint arXiv:1610.03112, 2016 – arxiv.org
… [12] Wojciech Zaremba, Ilya Sutskever, and Oriol Vinyals. Recurrent neural network regularization … Automatic recognition of conver- sational strategies in the service of a socially-aware dialog system. In 17th Annual SIGDIAL Meeting on Discourse and Dialogue, 2016 …

Spoken keyword detection using recurrent neural network language model
S Koike, A Lee – The Journal of the Acoustical Society of America, 2016 – asa.scitation.org
… Recurrent neural network language model (RNNLM) is introduced as linguistic constraint for both filler-filler and filler-keyword instead of N-gram, and experimental result on actual spoken data for a spoken dialogue system showed that our method can improve the keyword …

Tsung-Hsien (Shawn) Wen
P RNN, N LMs – mi.eng.cam.ac.uk
… Recurrent Neural Networks”, In NIPS workshop 2015, Machine Learning for SLU & Interaction. 6. Tsung-Hsien Wen, M. Gasic, N. Mrksic, P.-h. Su, D. Vandyke, S. Young, “Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems”, In …

Stylistic Transfer in Natural Language Generation Systems Using Recurrent Neural Networks
J Kabbara, JCK Cheung – EMNLP 2016, 2016 – aclweb.org
… Page 56. poses/directions. We plan to design similar recurrent neural network architectures to disentangle the style from the semantic content in text … 2015. Seman- tically conditioned lstm-based natural language gener- ation for spoken dialogue systems …

Recognizing emotions in spoken dialogue with hierarchically fused acoustic and lexical features
L Tian, J Moore, C Lai – Spoken Language Technology …, 2016 – ieeexplore.ieee.org
… emotional interaction modules in dialogue systems have been devel- oped with appraisal-based emotion models, and our goal is to build emotion … The LSTM model is a recurrent neural network with multiple hid- den layers and a special structure called “the memory cell” that …

Dialog state tracking for interview coaching using two-level LSTM
MH Su, CH Wu, KY Huang, TH Yang… – … (ISCSLP), 2016 10th …, 2016 – ieeexplore.ieee.org
… and their corresponding ten semantic slots illustrated in Table 2 are defined for the implementation of the interview dialog system … 4.2 LSTM for answer vector extraction Recently, many researchers have kept their eyes on recurrent neural networks (RNNs) which yield state-of …

Reading Comprehension using Entity-based Memory Network
X Wang, K Sudoh, M Nagata, T Shibata… – arXiv preprint arXiv …, 2016 – arxiv.org
… evaluation of gated recurrent neural networks on sequence modeling. arXiv preprint arXiv:1412.3555 (2014) 3. Dodge, J., Gane, A., Zhang, X., Bordes, A., Chopra, S., Miller, A., Szlam, A., Weston, J.: Evaluating prerequisite qualities for learning end-to-end dialog systems …

Sequential Match Network: A New Architecture for Multi-turn Response Selection in Retrieval-based Chatbots
Y Wu, W Wu, M Zhou, Z Li – arXiv preprint arXiv:1612.01627, 2016 – arxiv.org
… For each utterance-response pair, the model constructs a word-word similarity ma- trix and a sequence-sequence similarity matrix by the embedding of words and the hidden states of a recurrent neural network with gated unites (GRU) (Chung et al., 2014) respectively …

A Deep Learning Methodology for Semantic Utterance Classification in Virtual Human Dialogue Systems
D Datta, V Brashers, J Owen, C White… – … Conference on Intelligent …, 2016 – Springer
… the development of a deep learning methodology for semantic utterance classification (SUC) for use in domain-specific dialogue systems … in an utterance, we have proposed a shallow convolutional neural network (CNN) along with a recurrent neural network (RNN) that uses …

TEXT NORMALIZATION FOR AUTOMATIC SPEECH RECOGNITION SYSTEMS
AF VASILE, T BORO? – Editors: Maria Mitrofan Daniela Gîfu Dan Tufi? … – consilr.info.uaic.ro
… In Proceedings of the COLING/ACL on Main conference poster sessions, 33-40. Bang, J., Park, S., Lee, GG (2015). ASR Independent Hybrid Recurrent Neural Network Based Error Correction for Dialog System Applications. In Page 137 …

Using phone features to improve dialogue state tracking generalisation to unseen states
I Casanueva, T Hain, M Nicolao… – Proceeding of SIGDIAL …, 2016 – eprints.whiterose.ac.uk
… 1In a slot based dialogue system the dialogue states are defined as the set of possible value combinations for each slot … this paper, we propose a method to use ASR and phone-related general features to im- prove the generalisation of a Recurrent Neural Network (RNN) based …

Neural dialog state tracker for large ontologies by attention mechanism
Y Jang, J Ham, BJ Lee, Y Chang… – … Workshop (SLT), 2016 …, 2016 – ieeexplore.ieee.org
… Index Terms- Recurrent Neural Network, Dialog state tracking, DSTC5, Attention mechanism, Word embedding … Convolutional neural networks for multi-topic dialog state tracking,” in Pro- ceedings of the 7th International Workshop on Spoken Dialogue Systems (IWSDS), 2016 …

Sequence Generation & Dialogue Evaluation
R Lowe – cs.mcgill.ca
… networks.” ICLR, 2015. Rezende, DJ, Mohamed, S., & Wierstra, D. “Stochastic backpropagation and approximate inference in deep generative models.” ICML, 2014. Serban, Lowe, Charlin, Pineau. “A Survey of Available Corpora for Building Data-Driven Dialogue Systems.” …

Proceedings of the Workshop on Uphill Battles in Language Processing: Scaling Early Achievements to Robust Methods
A Louis, M Roth, B Webber, M White… – Proceedings of the …, 2016 – aclweb.org
… DialPort: A General Framework for Aggregating Dialog Systems Tiancheng Zhao, Kyusong Lee and Maxine Eskenazi … Stylistic Transfer in Natural Language Generation Systems Using Recurrent Neural Networks Jad Kabbara and Jackie Chi Kit Cheung …

LSTM ENCODER–DECODER FOR DIALOGUE RESPONSE GENERATION
Z Yu, C Yuan, X Wang, G Yang – workshop.colips.org
… Mei et al. [11] first encodes semantic via an LSTM-based recurrent neural network, then utilizes a novel coarse-to-fine … Furthermore, we focus on models which can exploit dialog histories to generate fluent, more human-like utterances for spoken dialogue systems. 3. MODEL …

Neural Networks for Natural Language Processing
L Mou – sei.pku.edu.cn
… [11] Yan Xu, Ran Jia, Lili Mou, Ge Li, Yunchuan Chen, Yangyang Lu, Zhi Jin. “Improved relation classification by deep recurrent neural networks with data … SequenceLevel Training ? Motivation: We don’t have the ground truth In a dialogue system, “The nature of of opendomain …

Knowledge Enhanced Hybrid Neural Network for Text Matching
Y Wu, W Wu, Z Li, M Zhou – arXiv preprint arXiv:1611.04684, 2016 – arxiv.org
… It captures sequential struc- tures of sentences in the pair by a Bidirectional Recurrent Neural Network with Gated units (BiGRU) (Bahdanau, Cho, and Bengio 2014), and constructs the similarity matrix with the hidden vectors given by BiGRU …

Quote Recommendation in Dialogue using Deep Neural Network
H Lee, Y Ahn, H Lee, S Ha, S Lee – … of the 39th International ACM SIGIR …, 2016 – dl.acm.org
… 3.1.1 Long short-term memory unit We use long short-term memory unit (LSTM) [3] which is a recurrent neural network with several gates which al- low networks to learn long-term … On-line policy optimisation of bayesian spoken dialogue systems via human interaction …

Designing Regularizers and Architectures for Recurrent Neural Networks
D Krueger – 2016 – papyrus.bib.umontreal.ca
Page 1. Université de Montréal Designing Regularizers and Architectures for Recurrent Neural Networks par David Krueger … tionist methods. Recurrent neural networks are a set of increasingly popular sequential models capable in principle of learning arbitrary algorithms …

Neural Discourse Modeling of Conversations
JM Pierre, M Butler, J Portnoff, L Aguilar – arXiv preprint arXiv:1607.04576, 2016 – arxiv.org
… While sequence to sequence models based on recurrent neural networks (RNNs) have shown initial promise in creating intelligible conversations [28], it has been noted that … The ubuntu dialogue corpus: A large dataset for research in unstructured multi-turn dialogue systems …

USTC at NTCIR-12 STC Task.
J Zhang, J Hou, S Zhang, LR Dai – NTCIR, 2016 – pdfs.semanticscholar.org
… To build a traditional dialogue system which contains several components[1], a lot of related technologies have been de- veloped such as dialogue state tracking[2], natural … The model encodes an input sentence with recurrent neural network and decodes an output sentence …

Hybrid Dialogue State Tracking for Real World Human-to-Human Dialogues.
K Sun, S Zhu, L Chen, S Yao, X Wu, K Yu – INTERSPEECH, 2016 – kaisun.org
… 8. References [1] S. Young, M. Gasic, B. Thomson, and JD Williams, “POMDP- based statistical spoken dialog systems: A review … http://www.aclweb.org/anthology/W/W13/W13-4073 [9] ——, “Word-based dialog state tracking with recurrent neural networks,” in Proceedings of the …

Recurrent Memory Addressing for describing videos
KK Agrawal, AK Jain, A Agarwalla, P Mitra – arXiv preprint arXiv …, 2016 – arxiv.org
… Recurrent Neural Networks (RNNs) with Long Short Term Memory (LSTM) [15] units or Gated Recur- rent Units (GRUs)[10], have similarly emerged as gener- ative models of choice for dealing with sequences in do- mains ranging from language modeling, machine transla- tion …

Multimodal Memory Modelling for Video Captioning
J Wang, W Wang, Y Huang, L Wang, T Tan – arXiv preprint arXiv …, 2016 – arxiv.org
… [42] exploit a hi- erarchical recurrent neural network decoder which … memory networks have been successfully applied to the tasks which need long-term dependency modelling, eg, textual ques- tion answering [3, 14], visual question answering [39] and dialog systems [8]. As we …

Proceedings of the Second International Afro-European Conference for Industrial Advancement AECIA 2015
A Abraham, K Wegrzyn-Wolska, AE Hassanien… – 2016 – books.google.com
… Windows….. 185 Hadhami Kaabi Solar Power Production Forecasting Based on Recurrent Neural Network….. 195 … Corral Page 14. Contents xv Dialogue Systems: Modeling and Prediction of their Dynamics….. 421 …

Crossmodal Language Grounding, Learning, and Teaching.
S Heinrich, C Weber, S Wermter, R Xie… – CoCo …, 2016 – pdfs.semanticscholar.org
… is still a major challenge: speech recognition is still limited to good signal-to-noise conditions or well adapted models; dialogue systems depend on a well … a The architecture consists of several neurocognitively plausible Continuous Time Recurrent Neural Networks (CTRNNs) …

Assisting Discussion Forum Users using Deep Recurrent Neural Networks
J Suorra Hagstedt P, O Mogren – Proceedings of the …, 2016 – publications.lib.chalmers.se
… 2015. Docu- ment modeling with gated recurrent neural network for sentiment classification. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 1422–1432 … 2013. Pomdp-based statistical spoken dialog systems: A review …

Mimicing the Man: a Persona-based Dialogue System
M Johnson – 2016 – stanford.edu
… this work, Serban, et al. compares variations of the HRED model with other state-of-the-art techniques in dialogue systems including baseline recurrent neural networks (RNN) as well as n-gram models. The group finds that the …

Deep Learning for Natural Language Processing-Research at Noah’s Ark Lab
H Li – 2016 – hangli-hl.com
… management Alan Turing Page 27. Natural Language Dialogue System – Retrieval based Approach index of messages and responses … matched responses Page 28. Retrieval based Dialogue System • Matching Models (Features) – Deep Match CNN – Deep Match Tree …

Challenges in Building Highly-Interactive Dialog Systems.
NG Ward, D DeVault – AI Magazine, 2016 – pdfs.semanticscholar.org
… Policy committee for adaptation in multi- domain spoken dialogue systems. In IEEE ASRU, 806–812. Geiger, JT; Eyben, F.; Schuller, B.; and Rigoll, G. 2013. Detecting overlapping speech with long short-term memory recurrent neural networks. In Interspeech, 1668–1672 …

Speech Intent Recognition for Robots
B Shen, D Inkpen – … and Computers in Sciences and in Industry …, 2016 – ieeexplore.ieee.org
… 2038-2042, October, 2008. [4] R. Jiang, TY Kee, H. Li, CY Wong, DK Limbu, “Development of Event-Driven Dialogue System for Social Mobile … [8] F. Raue, W. Byeon, TM Breuel, M. Liwicki, “Parallel sequence classification using recurrent neural networks and alignment”, 13th …

Sentence Level Recurrent Topic Model: Letting Topics Speak for Themselves
F Tian, B Gao, D He, TY Liu – arXiv preprint arXiv:1604.02038, 2016 – arxiv.org
… Different from conventional topic models that largely ignore the sequential order of words or their topic coher- ence, SLRTM gives full characterization to them by using a Recurrent Neural Networks (RNN) based framework …

Visual Fashion-Product Search at SK Planet
T Kim, S Kim, S Na, H Kim, M Kim, BK Jeon – arXiv preprint arXiv …, 2016 – arxiv.org
… 3431–3440, 2015. Mikolov, Tomas, Karafiát, Martin, Burget, Lukás, Cernocký, Jan, and Khudanpur, Sanjeev. Recurrent neural network based language model … Building end-to-end dialogue systems using generative hierarchical neural network models …

Optimising spoken dialogue systems using Gaussian process reinforcement learning for a large action set
TFW Nicholson, M Gaši? – pdfs.semanticscholar.org
… This is particularly important in dialog systems, where training data is either limited, or non-existent in which case live-interactions with real users … Recent work in [48] examined how to user recurrent neural networks (RNNs) to perform policy op- timisation in master action space …

Personified Autoresponder
A Mahendra – cs224d.stanford.edu
… 7 Conclusion Recurrent neural networks are quite powerful, with the right architecture these neural nets can learn complex tasks … [Ser+16] Iulian Vlad Serban et al. “Building End-To-End Dialogue Systems Using Generative Hierarchical Neural Network Models”. In: AAAI. 2016 …

A Dataset of Operator-client Dialogues Aligned with Database Queries for End-to-end Training
O Plátek, F Jur?í?ek – workshop.colips.org
… Plátek, O., B?lohlávek, P., Hude?ek, V., Jur?í?ek, F.: Recurrent neural networks for dialogue state tracking. arXiv preprint arXiv:1606.08733 (2016) 11. Raux, A., Langner, B., Bohus, D., Black, AW, Eskenazi, M.: Let’s go public! Taking a spoken dialog system to the real world …

Statistical Natural Language Generation
TH Wen, M Gasic – 2016 – mi.eng.cam.ac.uk
… On the difficulty of training recurrent neural networks. ICML 2013. • Tsung-Hsien Wen, Milica Gasic , Nikola Mrksic, Pei-Hao Su, David Vandyke, and Steve Young. Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems. In …

Addressee and Response Selection for Multi-Party Conversation.
H Ouchi, Y Tsuboi – EMNLP, 2016 – 2boy.org
… Our experiments on the dataset show that the recurrent neural network based models of our frameworks robustly pre- dict addressees and … Basically, the addressee detection has been tackled in the spoken/multimodal dialog system research, and the models largely rely on …

CONTEXT MEMORY NETWORKS FOR MULTI-OBJECTIVE SEMANTIC PARSING IN CONVERSATIONAL UNDERSTANDING
A Celikyilmaz, D Hakkani-Tur, G Tur, YN Chen, B Cao… – pdfs.semanticscholar.org
… Index Terms— recurrent neural networks, long-short term memory networks (LSTM), attention, embedding, mem- ory networks, spoken language understanding. 1. INTRODUCTION The Spoken Language Understanding (SLU) in conversa- tional dialog systems parses user …

Multi-label learning with the RNNs for Fashion Search
T Kim – 2016 – openreview.net
… Tomas Mikolov, Martin Karafiát, Lukás Burget, Jan Cernocký, and Sanjeev Khudanpur. Recurrent neural network based language model … Building end-to-end dialogue systems using generative hierarchical neural network models …

Non-sentential Question Resolution using Sequence to Sequence Learning.
V Kumar, S Joshi – COLING, 2016 – aclweb.org
… 2016. Building end- to-end dialogue systems using generative hierarchical neural network models. In AAAI. Ilya Sutskever, Oriol Vinyals, and Quoc V Le. 2014 … 2015. Translating videos to natural language using deep recurrent neural networks. In NAACL. Terry Winograd. 1971 …

The MSIIP system for dialog state tracking challenge 5
Y Su, M Li, J Wu – Spoken Language Technology Workshop …, 2016 – ieeexplore.ieee.org
… the application of deep learning has been used as well, like Deep Neural Network (DNN) [3], and Recurrent Neural Network (RNN) [4 … 2] Blaise Thomson and Steve Young, “Bayesian update of dialogue state: A pomdp framework for spoken dialogue systems,” Computer Speech …

A neural network approach for knowledge-driven response generation
P Vougiouklis, J Hare, E Simperl – 2016 – eprints.soton.ac.uk
… 2010. Recurrent neural network based language model … 2010. Vca: An experiment with a multiparty virtual chat agent. In Proceedings of the 2010 Workshop on Companionable Dialogue Systems, CDS ’10, pages 43–48, Stroudsburg, PA, USA …

NLU vs. Dialog Management: To Whom am I Speaking?
D Schnelle-Walka, S Radomski, B Milde… – Workshop on Smart … – researchgate.net
… 18. Mesnil, G., Dauphin, Y., Yao, K., Bengio, Y., Deng, L., Hakkani-Tur, D., He, X., Heck, L., Tur, G., Yu, D., et al. Using recurrent neural networks for slot filling in spoken language understanding … Research and commercial spoken dialog systems …

A Review on Deep Learning Algorithms for Speech and Facial Emotion Recognition
CP Latha, M Priya – APTIKOM Journal on Computer Science …, 2016 – jurnal.aptikom.or.id
… Some of the deep learning techniques like Convolutional Deep Neural Networks, Deep Boltzmann Machine, Deep Belief Networks, recurrent neural networks, deep neural networks and stacked auto encoders are applied to practical applications like pattern analysis, audio …

Optimising The Input Window Alignment in CD-DNN Based Phoneme Recognition for Low Latency Processing
AK Dhaka, G Salvi – arXiv preprint arXiv:1606.09163, 2016 – arxiv.org
… In dialogue systems, eg, long latencies may disrupt the natural turn- taking in the human-machine conversation … In [6] we used a hybrid of Recurrent Neural Networks (RNNs) and HMMs that was specifically designed for low latency process- ing …

Backchanneling via Twitter Data for Conversational Dialogue Systems
M Inaba, K Takahashi – International Conference on Speech and …, 2016 – Springer
… Experimental results demonstrated that our method can appropriately select backchannels to given inputs and significantly outperform baseline methods. Keywords. Conversational dialogue systems Recurrent neural network Backchanneling. 1 Introduction …

Ranking Responses Oriented to Conversational Relevance in Chat-bots.
B Wu, B Wang, H Xue – COLING, 2016 – aclweb.org
… 2014. Learning longer memory in recurrent neural networks. arXiv preprint arXiv:1412.7753. Alan Ritter, Colin Cherry, and William B Dolan … 2015. Building end-to- end dialogue systems using generative hierarchical neural network models. arXiv preprint arXiv:1507.04808 …

The splab at the NTCIR-12 Short Text Conversation Task.
K Wu, X Liu, K Yu – NTCIR, 2016 – pdfs.semanticscholar.org
… In the subtask, we build a single round of retrival-based dialogue system based on a repository of weibo data, which is provided by the competition’s organizers … Investigation of recurrent-neural-network architectures and learning methods for spoken language understanding …

Knowledge as a Teacher: Knowledge-Guided Structural Attention Networks
A Celikyilmaz, J Gao, L Deng – researchgate.net
… Abstract Natural language understanding (NLU) is a core component of a spoken dialogue system. Recently recurrent neural networks (RNN) ob- tained strong results on NLU due to their supe- rior ability of preserving sequential informa- tion over time …

Music Predictions Using Deep Learning. Could LSTM Networks be the New Standard for Collaborative Filtering?
E Keski-Seppälä, M Snellman – 2016 – diva-portal.org
… of a recurrent neural network (RNN) as a recommender system by comparing it to one of the most common recommender system implementations, the matrix factorization method … “Recurrent Neural Network” (RNN) är bättre på att förutspå vilken musik användare vill lyssna …

Improving the Probabilistic Framework for Representing Dialogue Systems with User Response Model.
M Li, Z Chen, J Wu – INTERSPEECH, 2016 – pdfs.semanticscholar.org
… Dialogue Management (DM) is the most important module in a spoken dialogue system … These meth- ods include Maximum Entropy Model [4, 5], Conditional Ran- dom Field [6, 7] and Recurrent Neural Network [8, 9]. Besides these statistical approaches, some robust domain …

Chatboj-Zizek: An Exercise in Automated Ideology
JY Engel, P Garzon – stanford.edu
… Some im- plementations, such as Google’s Neural Conversa- tional Model use recurrent neural networks using a sequence to sequence (seq2seq … In the rather entertainingly titled paper, ”How NOT To Evaluate Your Dialogue System: An Em- pirical Study of Unsupervised …

Neural Paraphrase Generation with Stacked Residual LSTM Networks
A Prakash, SA Hasan, K Lee, V Datla, A Qadir… – arXiv preprint arXiv …, 2016 – arxiv.org
… in the areas of machine translation (Cho et al., 2014; Bahdanau et al., 2015), speech recognition (Li and Wu, 2015), language modeling (Vinyals et al., 2015), and dialogue systems (Serban et al … Most of the deep learning models for NLP use Recurrent Neural Networks (RNNs) …

A survey of voice translation methodologies—Acoustic dialect decoder
H Krupakar, K Rajvel, B Bharathi… – Information …, 2016 – ieeexplore.ieee.org
… The advent of Recurrent Neural Networks (RNNs) into Natural Language Processing saw the field take a turn to an unexpected advancement … 1–6. [6] A. Graves, A. Mohamed, and G. Hinton, “Speech Recognition With Deep Recurrent Neural Networks,” Icassp, no. 3, pp …

Neural Network Approaches to Dialog Response Retrieval and Generation
L Nio, S Sakti, G Neubig, K Yoshino… – … on Information and …, 2016 – search.ieice.org
… unnatural re- sponses that are incomprehensible to the user [9]. There have been a number of works on response gener- ation for data-driven dialog systems … In this section, we explain about response generation and retrieval methods using LSTM recurrent neural networks …

ASR Error Management Using RNN Based Syllable Prediction for Spoken Dialog Applications
B Kim, J Choi, GG Lee – Advances in Parallel and Distributed Computing …, 2016 – Springer
… In: International workshop series on spoken dialogue systems technology (IWSDS). 4. Robinson T, Hochberg M, Renals S (1996) The use of recurrent neural networks in continuous speech recognition. In: Automatic speech and speaker recognition …

UT Dialogue System at NTCIR-12 STC.
S Sato, S Ishiwatari, N Yoshinaga, M Toyoda… – …, 2016 – pdfs.semanticscholar.org
… ABSTRACT This paper reports a dialogue system developed at the Uni- versity of Tokyo for participation in NTCIR-12 on the short … ranking of the chosen candidates in accordance with the perplexity given by Long Short-Term Memory-based Recurrent Neural Network (lstm-rnn …

Globally Coherent Text Generation with Neural Checklist Models
C Kiddon, L Zettlemoyer, Y Choi – … of the 2016 Conference on Empirical …, 2016 – aclweb.org
… Recurrent neural networks can generate lo- cally coherent text but often have difficulties representing what has already been generated … Evaluations on cooking recipes and dialogue system responses demonstrate high coherence with greatly improved semantic coverage of …

Nonparametric Bayesian Models for Spoken Language Understanding.
K Wakabayashi, J Takeuchi, K Funakoshi, M Nakano – EMNLP, 2016 – aclweb.org
Page 1. Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 2144–2152, Austin, Texas, November 1-5, 2016. c 2016 Association for Computational Linguistics Nonparametric …

Topic Augmented Neural Network for Short Text Conversation.
Y Wu, W Wu, Z Li, M Zhou – CoRR, 2016 – pdfs.semanticscholar.org
… (2011) employed statisti- cal machine translation techniques to generate re- sponses. Shang et al. (2015) formalized response generation as an encoding-decoding process, and learned generation models with recurrent neural networks …

1 Adaptability: Trainable end-to-end Natural Language Generation systems
I Konstas – ikonstas.net
… an in-order traversal of the graph, and feed every token in sequence through a multi-layer bidirectional Recurrent Neural Network (RNN); the … Existing dialogue systems currently fall under two extremes: (a) they are either confined to very small domains, eg, booking hotels or …

Hybrid of Evolutionary and Swarm Intelligence Algorithms for Prosody Modeling in Natural Speech Synthesis
M Sheikhan – International Journal of Information & Communication …, 2016 – journal.itrc.ac.ir
… context [8], emotional processing [9], speech to speech translation [10, 11], and automatic dialogue systems [12] … The parameters setting of BGSA is reported in Table 7. In addition, a standard Elman-type recurrent neural network [79] is used as a competitive prosody model …

Word representation using a deep neural network
Y Li – 2016 – search.proquest.com
… Fig 3.2. Recurrent Neural Network is a deep neural network in time (reproduced with permission from [47]) … The vectors can be used in various scenarios in NLP, such as machine translation, knowledge extraction, information retrieval, and dialogue systems …

Neural Document Embeddings for Intensive Care Patient Mortality Prediction
P Grnarova, F Schmidt, SL Hyland… – arXiv preprint arXiv …, 2016 – arxiv.org
… Following recent work in document classification [21] and dialogue systems [17], we adopt a two- layer architecture. Let d “ xs1,…,sny denote a patient’s record comprising n sentences … Recurrent neural networks for multivariate time series with missing values …

Proceedings of the 17th Annual Meeting of the Special Interest Group on Discourse and Dialogue
R Fernandez, W Minker, G Carenini… – Proceedings of the 17th …, 2016 – aclweb.org
… 13-15th, in close proximity to both INTERSPEECH 2016 and YRRSDS 2016, the Young Researchers’ Roundtable on Spoken Dialog Systems … 11 Joint Online Spoken Language Understanding and Language Modeling With Recurrent Neural Networks Bing Liu and Ian Lane …

An Analysis of Using Semantic Parsing for Speech Recognition
R Corona – cs.utexas.edu
… resources. It must be noted that in recent years there have been great advances in speech recognition which use deep recurrent neural networks (Graves & Jaitly, 2014; Xiong et al., 2016), particularly Long Short Term Memory networks (LSTMs) …

Shall I Be Your Chat Companion?: Towards an Online Human-Computer Conversation System
R Yan, Y Song, X Zhou, H Wu – … of the 25th ACM International on …, 2016 – dl.acm.org
… Neural Responding Machine (NRM) is developed as a generative way to respond to human messages, us- ing recurrent neural networks (RNN) [29] … A recurrent neural network (RNN) keeps a hidden state vector, which changes according to the input in each time step …

Generating Clinically Relevant Texts: A Case Study on Life-Changing Events.
M Oak, A Behera, T Thomas, CO Alm… – CLPsych@ HLT …, 2016 – pdfs.semanticscholar.org
… 4.1 Long-Short Term Memory Recurrent neural networks (RNN) are popular mod- els that have shown great potential in many natural language processing (NLP) tasks … 2015. Seman- tically conditioned lstm-based natural language gen- eration for spoken dialogue systems …

” Best Dinner Ever!!!”: Automatic Generation of Restaurant Reviews with LSTM-RNN
A Bartoli, A De Lorenzo, E Medvet… – Web Intelligence (WI …, 2016 – ieeexplore.ieee.org
… Methods for Natural Language Generation (NLG) are widely used in spoken dialogue systems [2], machine transla- tion [3], and image caption generation [4]. We are … A recent work [6] has shown the effectiveness of Recurrent Neural Networks (RNN) for NLG at character level …

Detecting Context Dependent Messages in a Conversational Environment
C Li, Y Wu, W Wu, C Xing, Z Li, M Zhou – arXiv preprint arXiv:1611.00483, 2016 – arxiv.org
… Differing from traditional dialogue systems (cf., (Young et al., 2013)) which rely on hand-crafted features and rules to generate reply sentences for specific applications such as voice dialling … We employ a Recurrent Neural Network (RNN) architecture to learn a message classifier …

Deep Recurrent Models with Fast-Forward Connections for Neural Machine Translation
JZYCX Wang, PLW Xu – pdfs.semanticscholar.org
… Moreover, NMT models can also be easily adapted to other tasks such as dialog systems (Vinyals and Le, 2015), question answering … Recurrent neural network (RNN) or its specific form LSTM is generally employed as the basic unit of the encoding and decoding function …

Recent Improvements on Error Detection for Automatic Speech Recognition.
Y Estève, S Ghannay, N Camelin – MMDA@ ECAI, 2016 – pdfs.semanticscholar.org
… and semantic features for theme identification in telephone conversations’, in 6th International Work- shop on Spoken Dialog Systems (IWSDS 2015 … [20] Yik-Cheung Tam, Yun Lei, Jing Zheng, and Wen Wang, ‘Asr error de- tection using recurrent neural network language model …

Online Sequence-to-Sequence Active Learning for Open-Domain Dialogue Generation.
N Asghar, P Poupart, X Jiang, H Li – arXiv preprint arXiv …, 2016 – researchgate.net
… 2016b. Mul- tiresolution recurrent neural networks: An applica- tion to dialogue response generation. arXiv preprint arXiv:1606.00776 … 2016. A network-based end-to-end trainable task-oriented dialogue system. arXiv preprint arXiv:1604.04562. [Weston2016] Jason Weston …

On Dialogue Breakdown: Annotation and Detection
K Funakoshi, R Higashinaka, M Inaba, Y Kobayashi… – workshop.colips.org
… errors and limited capabilities of machines, and are one of the major issues to be addressed in dialogue systems research … by use of NCM, LSTM, bag- of-words embedding, an extended NCM baseline CRF Word frequencies (RNN:Recurrent Neural Network, LSTM:Long Short …

Second International Afro-European Conference for Industrial Advancement AECIA 2015
A Abraham, K Wegrzyn-Wolska, AE Hassanien… – Springer
… Windows….. 185 Hadhami Kaabi Solar Power Production Forecasting Based on Recurrent Neural Network….. 195 … Contents Page 14. Dialogue Systems: Modeling and Prediction of their Dynamics….. 421 …

Invited Talks
G Rigoll, M Saraclar, E Dikici, E Arisoy, O Khomitsevich… – pdfs.semanticscholar.org
… 333 Gerasimos Arvanitis, Konstantinos Moustakas, and Nikos Fakotakis Recurrent Neural Networks for Hypotheses Re-Scoring . . . . . 341 Mikhail Kudinov Review of the Opus Codec in a WebRTC Scenario for Audio and Speech Communication …

Prototypical Recurrent Unit
D Long, R Zhang, Y Mao – arXiv preprint arXiv:1611.06530, 2016 – arxiv.org
… Abstract The difficulty in analyzing LSTM-like recurrent neural networks lies in the complex structure of the recurrent unit, which induces highly complex nonlinear dynamics. In this paper, we design a new simple recurrent unit, which we call Prototypical Recurrent Unit (PRU) …

A Study on Image Semantic Analysis Algorithm for Natural Language Understanding
J LUO, HUAJUN WANG, YANMEI LI… – Journal of Residuals …, 2016 – dpi-journals.com
… However, as an important research topic in this field, do we make full use of the great function of big data to build a data-driven natural language dialogue system? Here’s … Second, sort words by recurrent neural network. Each …

Evolvable dialogue state tracking for statistical dialogue management
K Yu, L Chen, K Sun, Q Xie, S Zhu – Frontiers of Computer Science, 2016 – Springer
… c Higher Education Press and Springer-Verlag Berlin Heidelberg 2015 Abstract Statistical dialogue management is the core of cognitive spoken dialogue systems (SDS) and has attracted great research interest … Dialogue management is the core of a dialogue system …

Human-Machine Interaction
M Heckmann, NB Thomsen, ZH Tan, B Lindberg… – pdfs.semanticscholar.org
Page 1. Contents Keynote …

Question Similarity Modeling with Bidirectional Long Short-Term Memory Neural Network
C An, J Huang, S Chang… – Data Science in …, 2016 – ieeexplore.ieee.org
… [9] Schuster M, Paliwal K K. Bidirectional recurrent neural networks[J]. Signal Processing, IEEE Transactions on, 1997, 45(11): 2673-2681 … Semantically conditioned lstm-based natural language generation for spoken dialogue systems[J]. arXiv preprint arXiv:1508.01745, 2015 …

Development of trainable policies for spoken dialogue systems
TC Le – dspace.cuni.cz
… neural network are also used and showing some improvements Zen and Sak [2015]. Some popular open-source libraries are Featival, MaryTTS and Flite. 1.2 Motivation From Section 1.1.3, we might see the importance of Dialogue Policy (DL) in Spoken Dialogue Systems (SDS …

Question Answering Using Deep Learning
E Stroh, P Mathur – pdfs.semanticscholar.org
… Recently, QA has also been used to develop dialog systems [1] and chatbots [2] designed to simulate human conversation … GRU and LSTM units allow recurrent neural networks (RNNs) to handle the longer texts required for QA …

Information Extraction from Spoken Diet Records
AK Gaddipati – ace.cs.ohiou.edu
… hamper the performance of the overall dialogue system. With the recent advancements in deep learning, Page 8. 8 neural systems that use Convolutional Neural Networks (CNN’s) or Recurrent Neural Networks (RNN’s) can …

Dead Man Tweeting
D Nilsson, M Sahlgren, J Karlgren – Workshop on Collecting and …, 2016 – diva-portal.org
… Mikolov, T., Karafiát, M., Burget, L., Cernocký, J., and Khudanpur, S. (2010). Recurrent neural network based language model … Semantically conditioned lstm- based natural language generation for spoken dialogue systems. In Proceedings of EMNLP, pages 1711–1721.

An Empirical Investigation of Word Clustering Techniques for Natural Language Understanding
DA Shunmugam, P Archana – International Journal of Engineering …, 2016 – ijesc.org
… A. Labeled Datasets The first set is internally collected multimedia data from live deployment scenarios of a spoken dialog system designed for entertainment search for Xbox One game console … Generic search utterances are sent by the dialog system to the Bing search engine …

Automatic Diacritics Restoration for Dialectal Arabic Text
AA Zayyan, M Elmahdy, H binti Husni… – International Journal of …, 2016 – ijcis.info
… [18] G. Abandah, A. Graves and B. Al-Shag, “Automatic diacritization of Arabic text using recurrent neural networks,” International Journal … From 2007 to 2011, he was pursuing his Ph.D. degree at the Dialogue Systems Group, Institute of Information Technology at the University …

A Corpus for Event Localization
C Ward – 2016 – bir.brandeis.edu
Page 1. A Corpus for Event Localization Master’s Thesis Presented to The Faculty of the Graduate School of Arts and Sciences Brandeis University Department of Computer Science Professor James Pustejovsky, Advisor In Partial Fulfillment of the Requirements for the Degree …

Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing
J Su, K Duh, X Carreras – Proceedings of the 2016 Conference on …, 2016 – aclweb.org
… Speech-based and Multimodal Approaches for Human versus Computer Addressee Detection Abstract: As dialog systems become ubiquitous, we must learn how to detect when a system is spoken to, and avoid mistaking human-human speech as computer-directed input …

Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
K Knight, A Nenkova, O Rambow – … of the 2016 Conference of the North …, 2016 – aclweb.org
… Abstractive Sentence Summarization with Attentive Recurrent Neural Networks Sumit Chopra, Michael Auli and Alexander M. Rush … Multi-domain Neural Network Language Generation for Spoken Dialogue Systems Tsung-Hsien Wen, Milica Gašic, Nikola Mrkšic, Lina M. Rojas …

Review of state-of-the-arts in artificial intelligence. Present and future of AI.
V Shakirov – alpha.sinp.msu.ru
… Recently, several state-of-the-arts were beaten with an algorithm that allows recurrent neural networks to learn how many computational steps to take between receiving an input and emitting an output[109](Mar2016) … ”Building End-To-End Dialogue Systems Using Generative …

Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
K Erk, NA Smith – Proceedings of the 54th Annual Meeting of the …, 2016 – aclanthology.info
Page 1. The 54th Annual Meeting of the Association for Computational Linguistics Proceedings of the Conference, Vol. 1 (Long Papers) August 7-12, 2016 Berlin, Germany Page 2. Platinum Sponsors Gold Sponsors Silver Sponsors ii Page 3. Bronze Sponsors …

Review of state-of-the-arts in artificial intelligence with application to AI safety problem
V Shakirov – arXiv preprint arXiv:1605.04232, 2016 – arxiv.org
… Recently, several state-of-the-arts were beaten with an algorithm that allows recurrent neural networks to learn how many computational steps to take between receiving an input and emitting an output[112](Mar2016) … ”Building End-To-End Dialogue Systems Using Generative …

An Exploratory Study on Process Representations
CN Naik – 2016 – search.proquest.com
… This chapter provides the background material and literature review for semantic role labeling and recurrent neural networks … in NLP and have been shown to benet question answering [7, 8], textual entailment [9], machine translation [1012], and dialogue systems [13, 14 …

Speech perception by humans and machines
MH Davis, O Scharenborg – Speech Perception and Spoken …, 2016 – books.google.com
… The most probable word sequence can then be transcribed, used to drive a dialogue system, or for other purposes (see Young, 1996 … H. Davis and Odette Scharenborg 1986), many researchers explored the possibility of using either static or recurrent neural networks in machine …

Personalized Natural Language Understanding.
X Liu, R Sarikaya, L Zhao, Y Ni, YC Pan – INTERSPEECH, 2016 – pdfs.semanticscholar.org
… Tracking for a Multi-Domain Dialog System using ASR Results,” Interspeech, Dresden Germany, September, 2015. [5] Puyang Xu and Ruhi Sarikaya, “Contextual domain classification in spoken language understanding systems using recurrent neural network”, Proceedings of …

MobileSSI: Asynchronous fusion for social signal interpretation in the wild
S Flutura, J Wagner, F Lingenfelser… – Proceedings of the 18th …, 2016 – dl.acm.org
… In PerCOM ’12 Workshops, pages 423–426. IEEE, 2012. [3] R. Brueckner and B. Schuller. Social Signal Classification using Deep BLSTM Recurrent Neural Networks. In ICASSP ’16, pages 4823–4827. IEEE, 2014. [4] C.-C. Chang and C.-J. Lin …

An empirical investigation of word class-based features for natural language understanding
A Celikyilmaz, R Sarikaya, M Jeong… – IEEE/ACM Transactions …, 2016 – dl.acm.org
… The first set is internally collected multimedia data from live deployment scenarios of a spoken dialog system designed for entertainment search for Xbox One game console … Generic search utterances are sent by the dialog system to the Bing search engine …

Real-Life Robustness
F Eyben – Real-time Speech and Music Classification by Large …, 2016 – Springer
… Yet, all of these systems do not use adaptive context learning as provided by Long Short-Term Memory Recurrent Neural Networks (LSTM-RNNs) (Sect … 2012) are used to overcome those limitations. Standard Recurrent Neural Networks (RNNs) (as applied by Gemello et al …

Deep learning in the automotive industry: Applications and tools
A Luckow, M Cook, N Ashcraft, E Weill… – Big Data (Big Data) …, 2016 – ieeexplore.ieee.org
… Voice dialog systems will become more natural and interactive with deep learning allowing a hands-free interaction with the vehicle … Conversely, recurrent neural networks add a hidden layer that is connected with itself for better speech recognition …

PersoNER: Persian Named-Entity Recognition
H Poostchi, E Zare Borzeshi, M Abdous… – The 26th International …, 2016 – opus.lib.uts.edu.au
… Applications Tim Baldwin Maria Liakata Dialog Processing and Dialog Systems, Multimodal Interfaces Nina Dethlefs Simon Keizer Giuseppe Riccardi Speech Recognition, Text-To-Speech, Spoken Language Understanding Florian Metze Chung-Hsien Wu …

Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers
Y Matsumoto, R Prasad – Proceedings of COLING 2016, the 26th …, 2016 – aclweb.org
… Applications Tim Baldwin Maria Liakata Dialog Processing and Dialog Systems, Multimodal Interfaces Nina Dethlefs Simon Keizer Giuseppe Riccardi Speech Recognition, Text-To-Speech, Spoken Language Understanding Florian Metze Chung-Hsien Wu …

Multi-behavioral Sequential Prediction for Collaborative Filtering
Q Liu, S Wu, L Wang – arXiv preprint arXiv:1608.07102, 2016 – pdfs.semanticscholar.org
… As two classical neural network methods for modeling sequences, recurrent neural networks can not well model short-term contexts, and the log-bilinear model is not suitable for long-term contexts … Fig. 2. Illustration of the Recurrent Neural Networks (RNN) model …

Gaussian Attention Model and Its Application to Knowledgebase Embedding and Question Answering
L Zhang, J Winn, R Tomioka – arXiv preprint arXiv:1611.02266, 2016 – arxiv.org
… Neelakantan et al. (2015a) proposed a method that combines relations to deal with compositional relations for knowledge base completion. Their key technical contribution is to use recurrent neural networks (RNNs) to encode a chain of relations …

On Generating Characteristic-rich Question Sets for QA Evaluation.
Y Su, H Sun, B Sadler, M Srivatsa, I Gur, Z Yan, X Yan – EMNLP, 2016 – aclweb.org
… Both based on single Freebase triples, SIMPLEQUES- TIONS (Bordes et al., 2015) employ human an- notators to formulate questions, while Serban et al. (2016) use a recurrent neural network to auto- matically formulate questions …

Scaling a Natural Language Generation System.
J Pfeil, S Ray – ACL (1), 2016 – aclweb.org
Page 1. Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, pages 1148–1157, Berlin, Germany, August 7-12, 2016. cO2016 Association for Computational Linguistics Scaling a Natural Language Generation System …

Dialog Management
M McTear, Z Callejas, D Griol – The Conversational Interface, 2016 – Springer
… parameters. The algorithms were evaluated on a statistical dialog system for the tourist information domain modeled as a POMDP. The … state). Wierstra et al. (2010) used recurrent neural networks (RNN) to approximate the policy. This …

Emotion Modelling via Speech Content and Prosody: In Computer Games and Elsewhere
B Schuller – Emotion in Games, 2016 – Springer
… approaches include a range of variants of Non-Negative Matrix Factorisation (NMF) [59] or the usage of (recurrent) neural networks (preferably with … synthesiser is often given in computer games, this may be a promising road in affect recognition for gaming or dialogue systems …

Spoken Language Understanding
M McTear, Z Callejas, D Griol – The Conversational Interface, 2016 – Springer
… In the examples above, there were utterances about setting alarms, finding restaurants, getting information about spoken dialog systems, and asking about personal feelings … Dialog state tracking involves estimating a user’s goal in a spoken dialog system …

N-gram Approximation of Latent Words Language Models for Domain Robust Automatic Speech Recognition
R Masumura, T Asami, T Oba, H Masataki… – … on Information and …, 2016 – search.ieice.org
Page 1. 2462 IEICE TRANS. INF. & SYST., VOL.E99–D, NO.10 OCTOBER 2016 PAPER Special Section on Recent Advances in Machine Learning for Spoken Language Processing N-gram Approximation of Latent Words Language Models for …

Computer-assisted pronunciation training: From pronunciation scoring towards spoken language learning
NF Chen, H Li – … Summit and Conference (APSIPA), 2016 Asia …, 2016 – ieeexplore.ieee.org
… Mattias Heldner, “An instantaneous vector representation of delta pitch for speaker-change prediction in conversation dialogue system,” in ICASSP … Klaus Zechner, Lei Chen, Jidong Tao, Aliaksei Ivanou, and Yao Qian, “Using bidirectional lstm recurrent neural networks to learn …

The Influence of Annotation, Corpus Design, and Evaluation on the Outcome of Automatic Classification of Human Emotions.
M Kächele, M Schels… – Front. ICT 3: 27. doi …, 2016 – pdfs.semanticscholar.org
… Subjects areasked topacka suitcase for a voyage to an unknown place using a voice controlled dialog system … Additional events are stimulated by setting malfunctions of the dialog system using external manipu- lations of the experimenter …

Institute of Communications Engineering Staff
M Bossert, R Fischer, W Minker, UC Fiebig… – Journal of Siberian …, 2016 – uni-ulm.de
… N. Rach, W. Minker and S. Ultes Towards an Argumentative Dialogue System Proceedings of 17th Workshop on Computational … L. Pragst, S. Ultes and W. Minker Recurrent Neural Network Interaction Quality Estimation Dialogues with Social Robots: Enablements, Analyses …

The Dark Side of NLP: Gefahren automatischer Sprachverarbeitung
M Strube – 2016 – pdfs.semanticscholar.org
… Johnston, Michael, Patrick Ehlen, Frederick G. Conrad, Michael F. Schober, Christopher An- toun, Stefanie Fail, Andrew Hupp, Lucas Vickers, Huiying Yan & Chan Zhang (2013). Spo- ken dialog systems for automated survey interviewing …

Institute of Information Technology
J Lindner, W Teich, A Linduska, M Mostafa… – Journal of Siberian …, 2016 – uni-ulm.de
… N. Rach, W. Minker and S. Ultes Towards an Argumentative Dialogue System Proceedings of 17th Workshop on Computational … L. Pragst, S. Ultes and W. Minker Recurrent Neural Network Interaction Quality Estimation Dialogues with Social Robots: Enablements, Analyses …

Semi-Supervised Learning with Sparse Autoencoders in Automatic Speech Recognition
AK Dhaka – 2016 – diva-portal.org
… 15 3.1.2 Recurrent Neural Networks … In dialogue systems, this tex- tual representation is fed to a language understanding module that extracts the semantic information to be handled by a dialogue manager …

Overview of NTCIR-12.
K Kishida, MP Kato – NTCIR, 2016 – research.nii.ac.jp
NTCIR (NII Testbeds and Community for Information access Research) Project …

Detecting Sarcasm in Multimodal Social Platforms
R Schifanella, P de Juan, J Tetreault… – Proceedings of the 2016 …, 2016 – dl.acm.org
Page 1. Detecting Sarcasm in Multimodal Social Platforms Rossano Schifanella University of Turin Corso Svizzera 185 10149, Turin, Italy schifane@di.unito.it Paloma de Juan Yahoo 229 West 43rd Street New York, NY 10036 pdjuan@yahoo-inc.com …

A learned representation for artistic style
V Dumoulin, J Shlens, M Kudlur, A Behboodi… – arXiv preprint arXiv …, 2016 – arxiv.org

Semi-supervised acoustic model training by discriminative data selection from multiple ASR systems’ hypotheses
S Li, Y Akita, T Kawahara – IEEE/ACM Transactions on Audio, Speech …, 2016 – dl.acm.org
Page 1. 1520 IEEE/ACM TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 24, NO. 9, SEPTEMBER 2016 Semi-Supervised Acoustic Model Training by Discriminative Data Selection From Multiple ASR Systems’ Hypotheses …

Personalised Dialogue Management for Users with Speech Disorders
I Casanueva – 2016 – etheses.whiterose.ac.uk
… MDP Markov Decision Process. PO Policy Optimization. POMDP Partially Observable Markov Decision Process. RL Reinforcement Learning. RNN Recurrent Neural Network. SDS Spoken Dialogue System. SLU Spoken Language Understanding. v Page 10. vi Acronyms …

Large Scale Data Enabled Evolution of Spoken Language Research and Applications
S Jothilakshmi, VN Gudivada – Handbook of Statistics, 2016 – Elsevier
Natural Language Processing (NLP) is an interdisciplinary field whose goal is to analyze and understand human languages. Natural languages are used in two forms.

Enabling collaborative geoinformation access and decision?making through a natural, multimodal interface
R Sharma, I Rauschert, H Wang – Citeseer
… 8 (query of spatial objects in a limited geographic region). This is arguably one of the most successful efforts thus far in verbal dialogue systems to databases (whether spatial or not), however, Glass and colleagues (Glass et al., 1995) reported performance of only …

Learning to Interpret and Generate Instructional Recipes
C Kiddon – 2016 – digital.lib.washington.edu
… plicit revision or generation of recipes. Laroche et al. (2013) proposed Cooking Coach, a spoken dialogue system to help a user search for recipes and prepare the recipe. A similar … in the way described above. Approaches based on recurrent neural network (RNN) archi …

Modeling dynamics of expressive body gestures in dyadic interactions
Z Yang, S Narayanan – IEEE Transactions on Affective …, 2016 – ieeexplore.ieee.org
… expressions and speech cues [20]. To consider the temporal dynamics of non- acted body gestures, a Recurrent Neural Network algorithm was employed for emotion recognition in the context of a video game [21]. In spite of the …

Power Data Classification: A Hybrid of a Novel Local Time Warping and LSTM
Y Li, H Hu, Y Wen, J Zhang – arXiv preprint arXiv:1608.04171, 2016 – arxiv.org
… as an upgrade of the recurrent neural network (RNN) [20]. RNN is used to handle sequential data with a special calculation process following the time step increment, while traditional neural network simply treats the sequence as a plain vector …

Energy-scalable speech recognition circuits
M Price – 2016 – dspace.mit.edu
… QFN quad flat no-leads ReL rectified linear RLE run-length encoding RNN recurrent neural network ROC receiver operating characteristic ROM read-only memory RTF real-time factor RTL register transfer level … In dialogue systems, the …

Emotion Identification from Spontaneous Communication
MK Dorry – 2016 – etd.aau.edu.et
Page 1. Addis Ababa University College of Natural Sciences Emotion Identification from Spontaneous Communication Mikiyas Kebede Dorry A Thesis Submitted to the Department of Computer Science in Partial Fulfilment for the Degree of Master of Science in …

Computer Vision and Natural Language Processing: Recent Approaches in Multimedia and Robotics
P Wiriyathammabhum, D Summers-Stay… – ACM Computing …, 2016 – dl.acm.org
… Recent methods use a CNN to detect visual features and using Recurrent Neural Networks (RNNs) [Karpathy and Fei-Fei 2015a] or Long-Short Term Memory (LSTM) [Vinyals et al. 2015] to generate the sentence description …

A review study of human-affection knowledge on usability engineering
D Lakshmi, R Ponnusamy – Advances in Human Machine …, 2016 – ieeexplore.ieee.org
… Introducing CURRENNT – the Munich Open-Source CUDA Recurrent Neural Network Toolkit … Sensors 2014, 14, 17491–17515. [50] Alonso-Martín, F.; Castro-González, A.; Luengo, F.; Salichs, M. Augmented Robotics Dialog System for Enhancing Human–Robot Interaction …

Speech Recognition Using Articulatory and Excitation Source
KS Rao, KE Manjunath – Springer
… Some of the topics covered in this series include the presentation of real life commercial deployment of spoken dialog systems, contemporary methods of speech parameterization, developments in information security for automated speech, forensic speaker recognition, use …

A Base Camp for Scaling AI
CJC Burges, T Hart, Z Yang, S Cucerzan… – arXiv preprint arXiv …, 2016 – arxiv.org
… An effective automated open domain dialog system will likely require a rich world model and the ability to perform commonsense reasoning over it [14] … 1We also use “TAL” to denote the overall dialog system itself, when the meaning is clear from context. 6 Page 7 …

Imitation learning for language generation from unaligned data
G Lampouras, A Vlachos – Proceedings of COLING 2016, the …, 2016 – eprints.whiterose.ac.uk
Page 1. Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 1101–1112, Osaka, Japan, December 11-17 2016. Imitation learning for language generation from unaligned data …

Fuel your imagination.
FN t BSUJDMF EPXOMPBET, BEFN t OFX – pdfs.semanticscholar.org
… Chen, and Zohreh Sojoudi 49 CONVERSATIONAL IN-VEHICLE DIALOG SYSTEMS Fuliang Weng, Pongtep Angkititrakul, Elizabeth E. Shriberg, Larry Heck, Stanley Peters, and John HL Hansen 61 RECENT ADVANCES IN …

Semi-supervised and unsupervised methods for categorizing posts in web discussion forums
K Perumal – arXiv preprint arXiv:1604.00119, 2016 – arxiv.org
… Other unsupervised techniques have been employed for the related tasks of dialogue act classification in spoken dialogue systems (Crook et al., 2009) and Twitter conversations (Ritter et al., 2010). Although they worked specifically on genres of text that are very …

Lexical and language modeling for Russian large vocabulary continuous speech recognition
S Zablotskiy – 2016 – d-nb.info
… directly or indirectly. First of all I would like to express my sincere gratitude to Prof. Wolfgang Minker for granting me the opportunity to work on my dissertation under his supervision in the Dialog Systems Group. I truly appreciate …

Character modeling through dialogue for expressive Natural Language Generation
GI Lin – 2016 – search.proquest.com
… 12. 3.1 Dialogue Corpus. 2015b], was created to build end-to-end dialogue systems with recurrent neural networks (RNN) and n-gram models. It contains dialogue of 3 turns between two interlocutors, which restricts the modeling data to dialogue with only two speakers …

Incremental Learning from Scratch Using Analogical Reasoning
V Letard, S Rosset, G Illouz – Tools with Artificial Intelligence …, 2016 – ieeexplore.ieee.org
Page 1. Incremental Learning From Scratch Using Analogical Reasoning Vincent Letard1,2,3, Sophie Rosset1 and Gabriel Illouz1,2,3 1LIMSI CNRS, France 2Université Paris Sud, France 3Université Paris Saclay, France firstname.lastname@limsi.fr …

Sign transition modeling and a scalable solution to continuous sign language recognition for real-world applications
K Li, Z Zhou, CH Lee – ACM Transactions on Accessible Computing …, 2016 – dl.acm.org
Page 1. 7 Sign Transition Modeling and a Scalable Solution to Continuous Sign Language Recognition for Real-World Applications KEHUANG LI, Georgia Institute of Technology ZHENGYU ZHOU, Research and Technology …

Language-independent methods for computer-assisted pronunciation training
A Lee – 2016 – dspace.mit.edu
… OOV Out of Vocabulary. PCA Principal Component Analysis. PCM Pulse-Code Modulation. ReLU Rectified Linear Unit. RNN Recurrent Neural Network. SGD Stochastic Gradient Descent. SGMM Subspace Gaussian Mixture Model. SVM Support Vector Machine …

Linguistic Knowledge in Data-Driven Natural Language Processing
Y Tsvetkov – 2016 – cs.cmu.edu
… and Persian, to grammatically parse Cantonese, to model Latin or Hebrew morphology, to build dialog systems for indigenous … Chapter 5 introduces “polyglot” language models, recurrent neural network models trained to predict symbol sequences in many different languages …

Exploiting Semantic and Topic Context to Improve Recognition of Proper Names in Diachronic Audio Documents
I Sheikh – 2016 – tel.archives-ouvertes.fr
Page 1. Exploiting Semantic and Topic Context to Improve Recognition of Proper Names in Diachronic Audio Documents Imran Sheikh To cite this version: Imran Sheikh. Exploiting Semantic and Topic Context to Improve Recognition …

Intelligent Robotics and Applications: 9th International Conference, ICIRA 2016, Tokyo, Japan, August 22-24, 2016, Proceedings
N Kubota, K Kiguchi, H Liu, T Obo – 2016 – books.google.com
… Human Action Recognition . . . . . 237 Zhi Chao Li, Ryad Chellali, and Yi Yang Page 18. Contents – Part I XIX A Study on Classification of Food Texture with Recurrent Neural Network. . . . Shuhei Okada, Hiroyuki …

Design and Evaluation of Statistical Parametric Techniques in Expressive Text-To-Speech: Emotion and Speaking Styles Transplantation
J Lorenzo Trueba – 2016 – oa.upm.es
… NAV Neutral Average Voice RNN Recurrent Neural Network SAT Speaker Adaptive Training … Applications such as virtual agents or avatars [Yaghoubzadeh et al., 2013; Krum et al., 2014; San-Segundo et al., 2012], dialog systems [Lutfi et al., 2013] are clear 1 Page 30 …

1st International Workshop on Multimodal Media Data Analytics (MMDA 2016)
S Vrochidis, M Melero, L Wanner, J Grivolla, Y Estève… – ecai2016.org
Page 1. ECAI 2016, MMDA 2016 workshop, August 2016 1st International Workshop on Multimodal Media Data Analytics (MMDA 2016) The rapid advancements of digital technologies, as well as the penetration of internet and …