Filters








3,028 Hits in 3.1 sec

Bootstrapping Conversational Agents with Weak Supervision

Neil Mallinar, Abhishek Shah, Rajendra Ugrani, Ayush Gupta, Manikandan Gurusankar, Tin Kam Ho, Q. Vera Liao, Yunfeng Zhang, Rachel K.E. Bellamy, Robert Yates, Chris Desmarais, Blake McGregor
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
In this paper, we present a framework called search, label, and propagate (SLP) for bootstrapping intents from existing chat logs using weak supervision.  ...  Many conversational agents in the market today follow a standard bot development framework which requires training intent classifiers to recognize user input.  ...  We also gathered user feedback to inform future work for the emerging applications of bootstrapping conversational agents, and more broadly training text classifiers, using weak supervision.  ... 
doi:10.1609/aaai.v33i01.33019528 fatcat:q6w76khzizccnkdjihn4pc5rwa

Bootstrapping Conversational Agents With Weak Supervision [article]

Neil Mallinar, Abhishek Shah, Rajendra Ugrani, Ayush Gupta, Manikandan Gurusankar, Tin Kam Ho, Q. Vera Liao, Yunfeng Zhang, Rachel K.E. Bellamy, Robert Yates, Chris Desmarais, Blake McGregor
2018 arXiv   pre-print
In this paper, we present a framework called search, label, and propagate (SLP) for bootstrapping intents from existing chat logs using weak supervision.  ...  Many conversational agents in the market today follow a standard bot development framework which requires training intent classifiers to recognize user input.  ...  We also gathered user feedback to inform future work for the emerging applications of bootstrapping conversational agents, and more broadly training text classifiers, using weak supervision.  ... 
arXiv:1812.06176v1 fatcat:fujpriekzrhh7jfgyz5spyp4ny

Unsupervised Learning of KB Queries in Task Oriented Dialogs [article]

Dinesh Raghu, Nikhil Gupta, Mausam
2020 arXiv   pre-print
Task-oriented dialog (TOD) systems converse with users to accomplish a specific task. This task requires the system to query a knowledge base (KB) and use the retrieved results to fulfil user needs.  ...  KB queries are usually annotated in real-world datasets and are learnt using supervised approaches to achieve acceptable task completion.  ...  We propose a novel cor-related attributes resilient RL (CARRL) approach for predicting KB queries using weak supervision to counter the effect of correlated attributes in the TOD KBs.  ... 
arXiv:2005.00123v1 fatcat:fd3zpk66kbeyvjbtbf2t5fftxu

Co-Adaptation of audio-visual speech and gesture classifiers

C. Mario Christoudias, Kate Saenko, Louis-Philippe Morency, Trevor Darrell
2006 Proceedings of the 8th international conference on Multimodal interfaces - ICMI '06  
Audio-Visual Agreement Recognition In this section, we apply multimodal co-training to the task of recognizing user agreement during multimodal interaction with a conversational agent.  ...  In this setting, the user interacts with an agent using speech and head gestures.  ... 
doi:10.1145/1180995.1181013 dblp:conf/icmi/ChristoudiasSMD06 fatcat:axyvbc3cwfgebcjaklaikyy5n4

Latent Intention Dialogue Models [article]

Tsung-Hsien Wen, Yishu Miao, Phil Blunsom, Steve Young
2017 arXiv   pre-print
Developing a dialogue agent that is capable of making autonomous decisions and communicating by natural language is one of the long-term goals of machine learning research.  ...  rely on hand-crafting a small state-action set for applying reinforcement learning that is not scalable or constructing deterministic models for learning dialogue sentences that fail to capture natural conversational  ...  ; secondly, the agent is capable of revising its conversational strategy based on an external reward within the same framework.  ... 
arXiv:1705.10229v1 fatcat:vvvdnld5dfgmhluru3aljb7d7q

Integrated Learning of Dialog Strategies and Semantic Parsing

Aishwarya Padmakumar, Jesse Thomason, Raymond J. Mooney
2017 Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers  
Previous research has used machine learning techniques to individually optimize these components, with different forms of direct and indirect supervision.  ...  an approach to integrate the learning of both a dialog strategy using reinforcement learning, and a semantic parser for robust natural language understanding, using only natural dialog interaction for supervision  ...  Acknowledgements We would like to thank the members of the UT Austin BWI group for several insightful discussions, and Subhashini Venugopalan for her help with revising the paper draft.  ... 
doi:10.18653/v1/e17-1052 dblp:conf/eacl/ThomasonPM17 fatcat:xkiltzzblzeenp5ughgvwpemyy

Less is More: Generating Grounded Navigation Instructions from Landmarks [article]

Su Wang, Ceslee Montgomery, Jordi Orbay, Vighnesh Birodkar, Aleksandra Faust, Izzeddin Gur, Natasha Jaques, Austin Waters, Jason Baldridge, Peter Anderson
2022 arXiv   pre-print
Using text parsers, weak supervision from RxR's pose traces, and a multilingual image-text encoder trained on 1.8b images, we identify 971k English, Hindi and Telugu landmark descriptions and ground them  ...  Generating such high-quality navigation instructions in novel environments is a step towards conversational navigation tools and could facilitate larger-scale training of instruction-following agents.  ...  and Beer Changpinyo for assistance with the multimodal mT5 implementation; and Igor Karpov, Ming Zhao and the Google ML Data Operations team for support collecting human evaluations.  ... 
arXiv:2111.12872v4 fatcat:ictdllrge5fcrhfqlztzkq6ovq

Diluted Near-Optimal Expert Demonstrations for Guiding Dialogue Stochastic Policy Optimisation [article]

Thibault Cordier, Tanguy Urvoy, Lina M. Rojas-Barahona, Fabrice Lefèvre
2020 arXiv   pre-print
A learning dialogue agent can infer its behaviour from interactions with the users. These interactions can be taken from either human-to-human or human-machine conversations.  ...  One solution to speedup the learning process is to guide the agent's exploration with the help of an expert.  ...  The team DATA-AI/NADIA at Orange-Labs earned unlimited access to the French HPC Jean Zay (IDRIS-CNRS) 2 with the project 10096 selected in the French contest Grands Challenges IA 2019.  ... 
arXiv:2012.04687v1 fatcat:uxj4bsznvfdhvfod2npdcehz24

Role of Relational Ties in the Relationship between Thriving at Work and Innovative Work Behavior: An Empirical Study

Riaz, Xu, Hussain
2019 European Journal of Investigation in Health, Psychology and Education  
Based on the socially embedded model of thriving, we aimed to assess the relevant related work on structured potential effects with relational ties (i.e., strong versus weak).  ...  Using partial least squares modeling on a sample of 412 observations (strong and weak ties), strong support was found for the theory-driven hypothesized relationships.  ...  The bootstrapping analysis showed that the indirect effect was significant with 0.1355, 0.0226, 5.997 .  ... 
doi:10.3390/ejihpe10010017 pmid:34542480 pmcid:PMC8314240 fatcat:lwlpayi3ljdk5dcisdzh4brjui

Fast and Scalable Expansion of Natural Language Understanding Functionality for Intelligent Agents

Anuj Kumar Goyal, Angeliki Metallinou, Spyros Matsoukas
2018 Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 3 (Industry Papers)  
We demonstrate that our proposed methods significantly increase accuracy in low resource settings and enable rapid development of accurate models with less data.  ...  Fast expansion of natural language functionality of intelligent virtual agents is critical for achieving engaging and informative interactions.  ...  NLU Functionality Expansion We focus on Amazon Alexa, an intelligent conversational agent that interacts with the user through voice commands and is able to process requests on a range of natural language  ... 
doi:10.18653/v1/n18-3018 dblp:conf/naacl/GoyalMM18 fatcat:jxta7bnu45d4dhhokxwaukqrbm

Large scale weakly and semi-supervised learning for low-resource video ASR [article]

Kritika Singh, Vimal Manohar, Alex Xiao, Sergey Edunov, Ross Girshick, Vitaliy Liptchinsky, Christian Fuegen, Yatharth Saraf, Geoffrey Zweig, Abdelrahman Mohamed
2020 arXiv   pre-print
Many semi- and weakly-supervised approaches have been investigated for overcoming the labeling cost of building high quality speech recognition systems.  ...  On the challenging task of transcribing social media videos in low-resource conditions, we conduct a large scale systematic comparison between two self-labeling methods on one hand, and weakly-supervised  ...  Introduction Recent advances in speech recognition systems have enabled successful large scale deployments of various customer-facing speech applications, e.g. personal conversational agents, automatic  ... 
arXiv:2005.07850v2 fatcat:l6zcwjqqijfdzh53qtm2goiv5y

Fast and Scalable Expansion of Natural Language Understanding Functionality for Intelligent Agents [article]

Anuj Goyal, Angeliki Metallinou, Spyros Matsoukas
2018 arXiv   pre-print
We demonstrate that our proposed methods significantly increase accuracy in low resource settings and enable rapid development of accurate models with less data.  ...  Fast expansion of natural language functionality of intelligent virtual agents is critical for achieving engaging and informative interactions.  ...  NLU Functionality Expansion We focus on Amazon Alexa, an intelligent conversational agent that interacts with the user through voice commands and is able to process requests on a range of natural language  ... 
arXiv:1805.01542v1 fatcat:lek7j373j5a6fjz6im245at74u

Large Scale Weakly and Semi-Supervised Learning for Low-Resource Video ASR

Kritika Singh, Vimal Manohar, Alex Xiao, Sergey Edunov, Ross Girshick, Vitaliy Liptchinsky, Christian Fuegen, Yatharth Saraf, Geoffrey Zweig, Abdelrahman Mohamed
2020 Interspeech 2020  
Many semi-and weakly-supervised approaches have been investigated for overcoming the labeling cost of building highquality speech recognition systems.  ...  On the challenging task of transcribing social media videos in low-resource conditions, we conduct a large scale systematic comparison between two self-labeling methods on one hand, and weakly-supervised  ...  Introduction Recent advances in speech recognition systems have enabled successful large scale deployments of various customer-facing speech applications, e.g. personal conversational agents, automatic  ... 
doi:10.21437/interspeech.2020-1917 dblp:conf/interspeech/SinghMXEGLFSZM20 fatcat:ujyynrud2vhk5g7geunmbay7dq

Time Series Anomaly Detection with label-free Model Selection [article]

Deokwoo Jung, Nandini Ramanan, Mehrnaz Amjadi, Sankeerth Rao Karingula, Jake Taylor, Claudionor Nunes Coelho Jr
2021 arXiv   pre-print
We develop a model variance metric that quantifies the sensitivity of anomaly probability with a bootstrapping method.  ...  In this paper, we propose LaF-AD, a novel anomaly detection algorithm with label-free model selection for unlabeled times-series data.  ...  Conversely, for more reliable anomaly prediction we havev m → 1 or 0 (i.e., µ m σ = 0).  ... 
arXiv:2106.07473v1 fatcat:tw3q5m7ofzbgxjkzxz55nqsqma

And That's A Fact: Distinguishing Factual and Emotional Argumentation in Online Dialogue

Shereen Oraby, Lena Reed, Ryan Compton, Ellen Riloff, Marilyn Walker, Steve Whittaker
2015 Proceedings of the 2nd Workshop on Argumentation Mining  
Using an annotated set of "factual" and "feeling" debate forum posts, we extract patterns that are highly correlated with factual and emotional arguments, and then apply a bootstrapping methodology to  ...  This process automatically produces a large set of patterns representing linguistic expressions that are highly correlated with factual and emotional language.  ...  We also present results for a supervised learner with bag-of-word features to assess the difficulty of this task.  ... 
doi:10.3115/v1/w15-0515 dblp:conf/naacl/OrabyRCRWW15 fatcat:3f4yuuxkvfcvbi5wekeyvkvnuq
« Previous Showing results 1 — 15 out of 3,028 results