Filters








1,420 Hits in 6.2 sec

Joint semantic utterance classification and slot filling with recursive neural networks

Daniel Guo, Gokhan Tur, Wen-tau Yih, Geoffrey Zweig
2014 2014 IEEE Spoken Language Technology Workshop (SLT)  
slot filling, in one jointly trained model.  ...  In this paper, we show that RecNNs can be used to perform the core spoken language understanding (SLU) tasks in a spoken dialog system, more specifically domain and intent determination, concurrently with  ...  , and in particular, we present a recursive network model to jointly perform the SLU tasks of domain detection, intent determination, and slot filling.  ... 
doi:10.1109/slt.2014.7078634 dblp:conf/slt/GuoTYZ14 fatcat:pi4f3u4skbafhckhlh6f72sixa

Intent Detection and Slot Filling with Capsule Net Architectures for a Romanian Home Assistant

Anda Stoica, Tibor Kadar, Camelia Lemnaru, Rodica Potolea, Mihaela Dînsoreanu
2021 Sensors  
We propose a customized capsule neural network architecture that performs intent detection and slot filling in a joint manner and we evaluate how well it handles utterances containing various levels of  ...  Variability in language when expressing one intent proves to be the biggest challenge encountered by the model.  ...  We propose a customized Capsule Neural Network architecture, SEMCAPSNET, which performs intent detection and slot filling jointly.  ... 
doi:10.3390/s21041230 pmid:33572405 fatcat:4zljkly7onas5epxdmbc7c466e

A survey of joint intent detection and slot-filling models in natural language understanding [article]

H. Weld, X. Huang, S. Long, J. Poon, S. C. Han
2021 arXiv   pre-print
and slot filling tasks.  ...  In this article, we describe trends, approaches, issues, data sets, evaluation metrics in intent classification and slot filling.  ...  With multi-turn dialogue [Bhargava et al. 2013 ] included the context from previous queries for the intent classification and slot filling of the current query.  ... 
arXiv:2101.08091v3 fatcat:ai6w2imilrfupf4m5fm2rjtzxi

Contextual domain classification in spoken language understanding systems using recurrent neural network

Puyang Xu, Ruhi Sarikaya
2014 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We evaluate our approach against SVM with and without contextual features.  ...  In a multi-domain, multi-turn spoken language understanding session, information from the history often greatly reduces the ambiguity of the current turn.  ...  SLU typically involves determining the user intent and extracting relevant semantic slots from the natural language sentence.  ... 
doi:10.1109/icassp.2014.6853573 dblp:conf/icassp/XuS14 fatcat:qbgurxt3gfhp7fglnaxibp2ody

Query Tracking for E-commerce Conversational Search: A Machine Comprehension Perspective [article]

Yunlun Yang, Yu Gong, Xi Chen
2018 arXiv   pre-print
quite different and more challenging due to more diverse user expressions and complex intentions.  ...  However, comparing with the natural language understanding in traditional task-oriented dialog which focuses on slot filling and tracking, the query understanding in E-commerce conversational search is  ...  Slot filling aims at extracting slot values and corresponding slots in input utterances.  ... 
arXiv:1810.03274v1 fatcat:idk6c4ey6zhqraxmonclqzw4oy

DR 4.4: Natural Multimodal Interaction Final Pro- totype

Bernd Kiefer, Ivana Kruijff-Korbayova, Anna Welker, Rifca Peters, Sarah McLeod
2019 Zenodo  
Furthermore, support for an experiment with different interaction styles, using modulated gestures, has been added, and the VOnDA compiler and run-time system has been heavily improved.  ...  , and for the Explainable AI module.  ...  This paper is dedicated to our colleague and friend Hans-Ulrich Krieger, the creator of HFC. Hope you found peace, wherever you are.  ... 
doi:10.5281/zenodo.3443669 fatcat:ez7jk76vmncshlzoivgqftx4ji

Neural Approaches to Conversational AI

Jianfeng Gao, Michel Galley, Lihong Li
2018 The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval - SIGIR '18  
us/research/publication/neural-approaches-toconversational-ai/ We thank Lihong Li, Bill Dolan and Yun-Nung (Vivian) Chen for contributing slides. 2  ...  Intent Classification 3. Slot Filling 51 RNN for Slot Tagging -I [Hakkani-Tur+ 16] • Variations: a. RNNs with LSTM cells b. Look-around LSTM c. Bi-directional LSTMs d.  ...  Passage The Panthers finished the regular season with a 15-1 record, and quarterback Cam Newton was named the 2015 NFL Most Valuable Player (MVP).  ... 
doi:10.1145/3209978.3210183 dblp:conf/sigir/GaoG018 fatcat:pnhrb5jgdfgnxac3hxy52a65pm

Kite

Toby Jia-Jun Li, Oriana Riva
2018 Proceedings of the 16th Annual International Conference on Mobile Systems, Applications, and Services - MobiSys '18  
A task model represents the logical backbone of a bot, on which Kite layers a question-answer interface generated using a hybrid rule-based and neural network approach.  ...  Recently, neural network models have been shown to be capable of generating natural "chitchat" conversations, but it is unclear whether they will ever work for task modeling.  ...  We also thank Chris Brockett and Michel Galley for their help in training the neural network transduction models, and Brad Myers for his feedback on drafts of this paper.  ... 
doi:10.1145/3210240.3210339 dblp:conf/mobisys/LiR18 fatcat:7aohxsk35vf6rbd3gykwtvsab4

Linguistically-Enriched and Context-Aware Zero-shot Slot Filling [article]

A.B. Siddique, Fuad Jamour, Vagelis Hristidis
2021 arXiv   pre-print
Thus, it is imperative that these models seamlessly adapt and fill slots from both seen and unseen domains -- unseen domains contain unseen slot types with no training data, and even seen slots in unseen  ...  We propose a new zero-shot slot filling neural model, LEONA, which works in three steps.  ...  The authors in [32] exploited regular expressions for few-shot slot filling, Prototypical Network was employed in [11] , and the authors in [20] extended the CRF model by introducing collapsed dependency  ... 
arXiv:2101.06514v1 fatcat:6xvgetmcwnburorej27h3kmv5q

Learning to Memorize in Neural Task-Oriented Dialogue Systems [article]

Chien-Sheng Wu
2019 arXiv   pre-print
In this thesis, we leverage the neural copy mechanism and memory-augmented neural networks (MANNs) to address existing challenge of neural task-oriented dialogue learning.  ...  Lastly, we tackle generation-based dialogue learning with two proposed models, the memory-to-sequence (Mem2Seq) and global-to-local memory pointer network (GLMP).  ...  corpus spanning over seven domains, containing 8438 multi-turn dialogues, with each dia- logue averaging 13.68 turns.  ... 
arXiv:1905.07687v1 fatcat:upvqqczyajdtnf45bcszxs6nai

Identifying Domain Independent Update Intents in Task Based Dialogs

Prakhar Biyani, Cem Akkaya, Kostas Tsioutsiouliklis
2018 Proceedings of the 19th Annual SIGdial Meeting on Discourse and Dialogue  
Given a user utterance, the intent of a slot-value pair is captured using dialog acts (DA) expressed in that utterance.  ...  We build a multi-class classification model using LSTM's to identify the type of UI in user utterances in the Restaurant and Shopping domains.  ...  et al., 2015) , a high-level neural networks API, with the Tensorflow (Abadi et al., 2015) backend.  ... 
doi:10.18653/v1/w18-5049 dblp:conf/sigdial/BiyaniAT18 fatcat:m5mpiqj2nferraeegu7zrygr2a

Towards End-to-End Learning for Dialog State Tracking and Management using Deep Reinforcement Learning

Tiancheng Zhao, Maxine Eskenazi
2016 Proceedings of the 17th Annual Meeting of the Special Interest Group on Discourse and Dialogue  
The model is able to interface with a relational database and jointly learn policies for both language understanding and dialog strategy.  ...  This paper presents an end-to-end framework for task-oriented dialog systems using a variant of Deep Recurrent Q-Networks (DRQN).  ...  The hyper-parameters of the neural network model are as follows: the size of turn embedding is 30; the size of LSTMs is 256; each policy network has a hidden layer of 128 with tanh activation.  ... 
doi:10.18653/v1/w16-3601 dblp:conf/sigdial/ZhaoE16 fatcat:aundy2dhmjdldndc2fveom7foy

Towards End-to-End Learning for Dialog State Tracking and Management using Deep Reinforcement Learning [article]

Tiancheng Zhao, Maxine Eskenazi
2016 arXiv   pre-print
The model is able to interface with a relational database and jointly learn policies for both language understanding and dialog strategy.  ...  This paper presents an end-to-end framework for task-oriented dialog systems using a variant of Deep Recurrent Q-Networks (DRQN).  ...  The opinions expressed in this paper do not necessarily reflect those of NSF. We would also like to thank Alan W Black for discussions on this paper.  ... 
arXiv:1606.02560v2 fatcat:i4akopezyzchzbonjfx57pecdi

Dual Learning for Semi-Supervised Natural Language Understanding [article]

Su Zhu, Ruisheng Cao, Kai Yu
2020 arXiv   pre-print
In this work, we introduce a dual task of NLU, semantic-to-sentence generation (SSG), and propose a new framework for semi-supervised NLU with the corresponding dual model.  ...  The framework is composed of dual pseudo-labeling and dual learning method, which enables an NLU model to make full use of data (labeled and unlabeled) through a closed-loop of the primal and dual tasks  ...  Intent Detection and Slot Filling in NLU Recently, motivated by a number of successful neural network and deep learning methods in natural language processing, many neural network architectures have been  ... 
arXiv:2004.12299v1 fatcat:ymineuzvhfdfroz7vgasqjzypm

An End-to-end Approach for Handling Unknown Slot Values in Dialogue State Tracking

Puyang Xu, Qi Hu
2018 Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
We describe in this paper an E2E architecture based on the pointer network (PtrNet) that can effectively extract unknown slot values while still obtains state-of-the-art accuracy on the standard DSTC2  ...  We also provide extensive empirical evidence to show that tracking unknown values can be challenging and our approach can bring significant improvement with the help of an effective feature dropout technique  ...  In the absence of SLU providing fine-grained semantic features, the E2E approaches these days typically rely on variants of neural networks such as recurrent neural networks (RNN) or memory networks (  ... 
doi:10.18653/v1/p18-1134 dblp:conf/acl/XuH18 fatcat:gokkwd5mqbedhgeouwhrbslhha
« Previous Showing results 1 — 15 out of 1,420 results