28,582 Hits in 4.1 sec

Playing with Embeddings : Evaluating embeddings for Robot Language Learning through MUD Games

Anmol Gulati, Kumar Krishna Agrawal
2017 Proceedings of the 2nd Workshop on Evaluating Vector Space Representations for NLP  
In this paper, we study the suitability of using general purpose word-embeddings for language learning in robots.  ...  Based in a risk-reward setting, we review the effectiveness of the embeddings in navigating tasks in fantasy games, as an approximation to their performance on more complex scenarios, like language assisted  ...  In a similar spirit, MUD games provide a rich playground for define complex scenarios solely with textual descriptions.  ... 
doi:10.18653/v1/w17-5305 dblp:conf/repeval/GulatiA17 fatcat:36e7ovq7w5fmfp4ojuxbivxa24

Deep Reinforcement Learning-based Text Anonymization against Private-Attribute Inference

Ahmadreza Mosallanezhad, Ghazaleh Beigi, Huan Liu
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
User-generated textual data is rich in content and has been used in many user behavioral modeling tasks.  ...  Our approach first extracts a latent representation of the original text w.r.t. a given task, then leverages deep reinforcement learning to automatically learn an optimal strategy for manipulating text  ...  Acknowledgements The authors would like to thank Jundong Li for his help throughout the paper.  ... 
doi:10.18653/v1/d19-1240 dblp:conf/emnlp/MosallanezhadBL19 fatcat:wld7k4kt2vhe3h6pmf76u7ftny

The Road to Know-Where: An Object-and-Room Informed Sequential BERT for Indoor Vision-Language Navigation [article]

Yuankai Qi, Zizheng Pan, Yicong Hong, Ming-Hsuan Yang, Anton van den Hengel, Qi Wu
2021 arXiv   pre-print
Our sequential BERT also enables the visual-textual clues to be interpreted in light of the temporal context, which is crucial to multi-round VLN tasks.  ...  Vision-and-Language Navigation (VLN) requires an agent to find a path to a remote location on the basis of natural-language instructions and a set of photo-realistic panoramas.  ...  However, most of them are designed for one-round decision tasks (e.g., VQA).  ... 
arXiv:2104.04167v2 fatcat:cq34ceywgrbdlhhzkuygkde764

Joint Multimodal Embedding and Backtracking Search in Vision-and-Language Navigation

Jisu Hwang, Incheol Kim
2021 Sensors  
JMEBS uses both multimodal context and temporal context.  ...  This study proposes a novel deep neural network model (JMEBS), with joint multimodal embedding and backtracking search for VLN tasks.  ...  In summary, our contributions include: • A novel deep neural network model, with joint multimodal embedding and backtracking search for VLN tasks. • Rich context utilization based on a transformer-based  ... 
doi:10.3390/s21031012 pmid:33540789 pmcid:PMC7867342 fatcat:wgtyeowcrnbdjpq2b2jcufglhq

A theory of multiformat communication: mechanisms, dynamics, and strategies

Jordan W. Moffett, Judith Anne Garretson Folse, Robert W. Palmatier
2020 Journal of the Academy of Marketing Science  
and to identify any gaps (e.g., AI agents, simulated cues).  ...  The authors then review bilateral communication research in light of the identified theoretical gaps, to inform their framework.  ...  for the exchange. • Mechanical and analytical (i.e., lower intelligence) tasks are easier and more efficient for AI agents.  ... 
doi:10.1007/s11747-020-00750-2 pmid:33199929 pmcid:PMC7658432 fatcat:z2ltc2wrencwflwx2hg5wb5wq4

A Survey of Reinforcement Learning Informed by Natural Language

Jelena Luketina, Nantas Nardelli, Gregory Farquhar, Jakob Foerster, Jacob Andreas, Edward Grefenstette, Shimon Whiteson, Tim Rocktäschel
2019 Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence  
Finally, we call for the development of new environments as well as further investigation into the potential uses of recent Natural Language Processing (NLP) techniques for such tasks.  ...  We survey the state of the field, including work on instruction following, text games, and learning from textual domain knowledge.  ...  Information contained in both generic and task-specific large textual corpora may be highly valuable for decision making.  ... 
doi:10.24963/ijcai.2019/880 dblp:conf/ijcai/LuketinaNFFAGWR19 fatcat:vyjt3t4kbjbkdgyb2ra3quppbq

Refocusing on Relevance: Personalization in NLG [article]

Shiran Dudy, Steven Bedrick, Bonnie Webber
2021 arXiv   pre-print
of as a crucial tool for designing user-oriented text-generating tasks.  ...  In this work, we argue that NLG systems in general should place a much higher level of emphasis on making use of additional context, and suggest that relevance (as used in Information Retrieval) be thought  ...  This research was supported by the NSF National AI Institute for Student-AI Teaming (iSAT) under grant DRL 2019805, as well as the National Institute On Deafness And Other Communication Disorders of the  ... 
arXiv:2109.05140v1 fatcat:pqghixqbq5grlbrootumsgbr4u

Image-Grounded Conversations: Multimodal Context for Natural Question and Response Generation [article]

Nasrin Mostafazadeh, Chris Brockett, Bill Dolan, Michel Galley, Jianfeng Gao, Georgios P. Spithourakis, Lucy Vanderwende
2017 arXiv   pre-print
Experiments with models trained on social media data show that the combination of visual and textual context enhances the quality of generated conversational turns.  ...  We present a novel task, Image-Grounded Conversations (IGC), in which natural-sounding conversations are generated about a shared image.  ...  Moreover, Q F N =I F N for 32% our annotations, whereas Q F N =T F N for 47% of the triplets, again, showing the effectiveness of rich textual context in determining the question to be asked.  ... 
arXiv:1701.08251v2 fatcat:psa5hxyiefaszlnkjbgx2mjxey

Emotional Dialogue Generation using Image-Grounded Language Models

Bernd Huber, Daniel McDuff, Chris Brockett, Michel Galley, Bill Dolan
2018 Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems - CHI '18  
Humans rely on much more than language during conversations; visual information is key to providing context.  ...  We show that key qualities of the generated dialogue can be manipulated by the features used for training the agent.  ...  Designing Conversational Agents Our work has important implications for the design of conversational agents.  ... 
doi:10.1145/3173574.3173851 dblp:conf/chi/HuberMBGD18 fatcat:splo6zphs5dwlanlhmro6b2q3e

From domain modeling to collaborative domain construction

Gerhard Fischer, Stefanie Lindstaedt, Jonathan Ostwald, Markus Stolze, Tamara Sumner, Beatrix Zimmermann
1995 Proceedings of the conference on Designing interactive systems processes, practices, methods, & techniques - DIS '95  
Domain-oriented systems offer many potential benefits for end-users such as more intuitive interfaces, better task support, and knowledge-based assistance.  ...  While developing a system for service representatives at NYNEX, we ot.served how informal communication takes place.  ...  We thank Andrea s d Girgensohn for his work with Expectation Agents.  ... 
doi:10.1145/225434.225443 dblp:conf/ACMdis/FischerLOSSZ95 fatcat:dykbn6mhmjbf3ljmrou5tc6jcu

Managing cognitive load in educational multi-user virtual environments: reflection on design practice

Brian C. Nelson, Benjamin E. Erlandson
2007 Educational technology research and development  
In this paper, we explore how the application of multimedia design principles may inform the development of educational multi-user virtual environments (MUVEs).  ...  Our discussion describes a blueprint for research implementations that we are undertaking to systematically investigate the effect of an educational MUVE interface design based on multimedia principles-implementations  ...  On one hand, virtual worlds can be created that contain a rich set of visual and textual elements necessary to recreate a real-world context for exploring problems and developing hypotheses.  ... 
doi:10.1007/s11423-007-9082-1 fatcat:uqw2b2yscbgytdkddy73pert7q

Practical language-based editing for software engineers [chapter]

Michael L. Van De Vanter
1995 Lecture Notes in Computer Science  
Solutions require better understanding of software engineers and their tasks, appropriate design metaphors, new architectural organizations, and design for adaptation and extension.  ...  Major usability requirements for such systems include familiar, unrestricted text editing; coherent user interaction with software; rich, dynamic information display; multiple alternative views; uninterrupted  ...  Design Metaphors The requirements of Section 3 demand a system rich in functionality.  ... 
doi:10.1007/bfb0035821 fatcat:xb4a63t6ubgrrhxb5ugmxlogca

Making Topic-Specific Report and Multimodal Presentation Automatically by Mining the Web Resources

Shaikh Mostafa Al Masum, Mitsuru Ishizuka
2006 2006 IEEE/WIC/ACM International Conference on Web Intelligence (WI 2006 Main Conference Proceedings)(WI'06)  
So we set the mood for this text to disgust to the agent who speaks-out this sentence.  ...  Due to availability and accessibility of enormous internet-based resources and dynamic nature of web pages, the task of information retrieval is becoming more challenging and gradually tricky.  ...  [4] [10]), Web queries, Wrapper Induction for Information, Effective Resource Discovery, User Preference-based search and Content or Context based summarization.  ... 
doi:10.1109/wi.2006.113 dblp:conf/webi/MasumI06 fatcat:lnycqxvh6bgdljjcd2n362k2iq

Towards human-agent knowledge fusion (HAKF) in support of distributed coalition teams [article]

Dave Braines, Federico Cerutti, Marc Roig Vilamala, Mani Srivastava, Lance Kaplan Alun Preece, Gavin Pearson
2020 arXiv   pre-print
Future coalition operations can be substantially augmented through agile teaming between human and machine agents, but in a coalition context these agents may be unfamiliar to the human users and expected  ...  The human agents are also able to bring their local knowledge to the team, observing the situation unfolding and deciding which key information should be communicated to the machine agents to enable them  ...  agent consumability, and through negotiation can form the basis for coalition interoperability through fine-grained sharing of information from this graph according to policies and agent-managed service  ... 
arXiv:2010.12327v1 fatcat:jukoa5qcbvb5bkx2zymiybttxa

A Semantic Case-Based Reasoning Framework for Text Categorization [chapter]

Valentina Ceausu, Sylvie Desprès
2007 Lecture Notes in Computer Science  
This paper presents a semantic case-based reasoning framework for text categorization. Text categorization is the task of classifying text documents under predefined categories.  ...  We propose a textual case-based reasoning approach (TCBR), which allows us to integrate both textual and domain knowledge aspects in order to carry out this categorization.  ...  Thus, methods, like in particular Information Retrieval, are not sufficient to create knowledge-rich case representations from text.  ... 
doi:10.1007/978-3-540-76298-0_53 fatcat:bfop7sb7o5gj3gs6h7khpuprnu
« Previous Showing results 1 — 15 out of 28,582 results