Filters








83,998 Hits in 6.7 sec

Learning Content Selection Rules for Generating Object Descriptions in Dialogue

P. W. Jordan, M. A. Walker
2005 The Journal of Artificial Intelligence Research  
To our knowledge, this is the first reported experiment of a trainable content selection component for object description generation in dialogue.  ...  model, and Jordan's (2000b) intentional influences model, and use these feature sets in a machine learning experiment to automatically learn a model of content selection for object descriptions.  ...  Experimental Method The experiments utilize the rule learning program ripper (Cohen, 1996) to learn the content selection component of an object description generator from the object descriptions in  ... 
doi:10.1613/jair.1591 fatcat:bb2oqq3pgreelfjai5vqvqyyye

Get Involved in an Interactive Virtual Tour of Brest Harbour: Follow the Guide and Participate [chapter]

Mukesh Barange, Pierre De Loor, Vincent Louis, Ronan Querrec, Julien Soler, Thanh-Hai Trinh, Éric Maisel, Pierre Chevaillier
2011 Lecture Notes in Computer Science  
We applied these principles to develop Brest'Coz, an interactive virtual tour for the learning of shipbuilding techniques used in France in early 18 th century.  ...  We propose to use the semantic modelling of the VE and the activities of the agents for the conceptualisation of the dialogue. Meta level semantic information are used as arguments in NLU/NLG rules.  ...  This work was supported in part by a grant from Région Bretagne and in part by the Anr (Corvette project Anr-10-Cord-012).  ... 
doi:10.1007/978-3-642-23974-8_10 fatcat:fogxyz6kzrhlrist35sbsdv4ta

Discourse Coherence, Reference Grounding and Goal Oriented Dialogue [article]

Baber Khalid, Malihe Alikhani, Michael Fellner, Brian McMahan, Matthew Stone
2020 arXiv   pre-print
As first steps towards implementing the approach, we describe a simple dialogue system in a referential communication domain that accumulates constraints across discourse, interprets them using a learned  ...  In this paper, we argue for a new approach, inspired by coherence-based models of discourse such as SDRT , in which utterances attach to an evolving discourse structure and the associated knowledge graph  ...  We thank the Semdial reviewers for helpful comments, and the Mechanical Turk annotators for their contributions.  ... 
arXiv:2007.04428v1 fatcat:fmxomllhqbci3junqf2gd66kxa

Conditional Text Generation for Harmonious Human-Machine Interaction [article]

Bin Guo, Hao Wang, Yasan Ding, Wei Wu, Shaoyang Hao, Yueqi Sun, Zhiwen Yu
2020 arXiv   pre-print
In recent years, with the development of deep learning, text generation technology has undergone great changes and provided many kinds of services for human beings, such as restaurant reservation and daily  ...  We further make an investigation of existing CTG fields and propose several general learning models for CTG. Finally, we discuss the open issues and promising research directions of CTG.  ...  open research issues and promising research directions should be studied, such as long text generation, multimodal data translation, and lifelong learning.  ... 
arXiv:1909.03409v2 fatcat:s2zfmwxtubgwjks4luoby6vdoq

Developing inclusive e-learning systems

Anthony Savidis, Dimitris Grammenos, Constantine Stephanidis
2006 Universal Access in the Information Society  
Inclusive e-learning is the outcome from the application of e-inclusion design and implementation methods in the context of e-learning systems.  ...  In this context, the primary emphasis is placed on the reporting of the design and implementation aspects to accommodate the inclusive system characteristics, rather than on the typical e-learning software  ...  In other words, learning objects can be polymorphic content entities.  ... 
doi:10.1007/s10209-006-0024-1 fatcat:ps5yko35ergezgbc3c5v5apdke

Applying CBR to manage argumentation in MAS

Stella Heras, Vicente Julian, Vicente Botti
2010 International Journal of Reasoning-based Intelligent Systems  
In those settings, agents must be able to generate, select and send arguments to other agents that will evaluate them in their turn.  ...  This paper analyses the advantages of applying case-based reasoning to manage arguments in multi-agent systems dialogues, identifies open issues and proposes new ideas to tackle them.  ...  A justified prediction can be generated using CBR in combination with any learning method, like decision trees or Lazy Induction of Descriptions (LID) (Armengol and Plaza, 2001) .  ... 
doi:10.1504/ijris.2010.034906 fatcat:sgqmggcrj5g35ecus3l5o5gg2m

Adaptive Generation in Dialogue Systems Using Dynamic User Modeling

Srinivasan Janarthanam, Oliver Lemon
2014 Computational Linguistics  
We also show that adaptation to users' domain knowledge results in improving task success (99.47% for the learned policy vs. 84.7% for a hand-coded baseline) and reducing dialogue time of the conversation  ...  We discuss the overall advantages of this method and how it can be extended to other levels of adaptation such as content selection and dialogue management, and to other domains where adapting to users  ...  to content selection and dialogue management.  ... 
doi:10.1162/coli_a_00203 fatcat:3v3wetxiabd3zhtmta3zsdflvq

Natural Language Processing in Serious Games: A state of the art

Davide Picca, Dominique Jaccard, Gérald Eberlé
2015 International Journal of Serious Games  
Support vector machines Support vector machines are supervised learning methods generally used for classification tasks.  ...  SVM are used in DeLearyous and Mission Rehearsal for classification tasks.  ...  The dialogue management system also interacts with a task planner, an action selector and an emotion model to generate a content selection.  ... 
doi:10.17083/ijsg.v2i3.87 fatcat:smpb6w3bf5fsvhjgkcskxkavmm

Challenges for a CBR framework for argumentation in open MAS

Stella Heras, Vicente Botti, Vicente Julián
2009 Knowledge engineering review (Print)  
Some approaches that apply CBR to manage argumentation in MAS have been proposed in the literature. These improve agents' argumentation skills by allowing them to reason and learn from experiences.  ...  The high dynamism of open MAS gives rise to potential conflicts between agents and thus, to a need for a mechanism to reach agreements.  ...  The cases contents store descriptions that characterise the agents' context in a previous negotiation.  ... 
doi:10.1017/s0269888909990178 fatcat:ilsqei4kfbcknalukjnin6z3k4

Learning what to say and how to say it: Joint optimisation of spoken dialogue management and natural language generation

Oliver Lemon
2011 Computer Speech and Language  
For content selection, recent research has shown that given a corpus of texts and the database of facts or events it describes, content selection rules can be learned [1, 5] .  ...  However, in practice, the content planning and sentence planning components usually consist of domain-specific rules, or general rules that are tuned for the specific domain.  ... 
doi:10.1016/j.csl.2010.04.005 fatcat:4afk2pvwjbabrntpkgezdy5dae

Spoken dialogue processing method using inductive learning with genetic algorithm

Yasutomo Kimura, Kenji Araki, Yoshio Momouchi, Koji Tochinai
2004 Systems and Computers in Japan  
Most of the spoken dialogue systems up to the present are task-oriented, where the processing is based on the prespecified generation rules and database.  ...  This paper describes a spoken dialogue processing, which includes the learning using spoken dialogue examples.  ...  In inductive learning, the rule pair with only one different part is considered as the object for learning.  ... 
doi:10.1002/scj.10204 fatcat:ptzz22jpm5gpzahxxsrg3q6zvq

A case-based architecture for a dialogue manager for information-seeking processes

Anne Tissen
1991 Proceedings of the 14th annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR '91  
In this paper, we propose a case-based architecture for a dialogue manager.  ...  In our approach, this functionality will be provided by the dialogue manager operating on a set of case-based dialogue plans.  ...  In our first version, we use a set of heuristic rules to select a plan horn the plan memory.  ... 
doi:10.1145/122860.122875 dblp:conf/sigir/Tissen91 fatcat:vyhasvwkmngpnkrd7pvyahdgma

A conversation acts model for generating spoken dialogue contributions

Amanda J Stent
2002 Computer Speech and Language  
It has been implemented in the TRIPS spoken dialogue system, and includes components that plan content, perform surface generation for different modalities, and coordinate output production.  ...  In this paper, we describe a generation system for spoken dialogue that not only produces coherent, informative and responsive dialogue contributions, but also explicitly models human styles of interaction  ...  An architecture for a generic dialogue shell. Natural Language Engineering special issue on Best Practice in Spoken Language Dialogue Systems Engineering 6(3). 350 A.J. Stent  ... 
doi:10.1016/s0885-2308(02)00009-8 fatcat:65c4kgv6yjgwblny4euekeboqe

Designing a toolkit to support dialogue in learning

John Cook, Martin Oliver
2002 Computers & Education  
This is then exemplified with the specific case of the design of a toolkit for using dialogue in learning.  ...  This paper describes one strategy for addressing this issue: the development of a toolkit that supports the process of planning and reflection that practitioners must engage in when attempting to use dialogue  ...  Acknowledgements Our thanks to colleagues who took part in the study and to the reviewers for their helpful comments.  ... 
doi:10.1016/s0360-1315(01)00076-8 fatcat:dgqylason5cd7lbu3qqs5lb6jy

Learning to Speak and Act in a Fantasy Text Adventure Game [article]

Jack Urbanek, Angela Fan, Siddharth Karamcheti, Saachi Jain, Samuel Humeau, Emily Dinan, Tim Rocktäschel, Douwe Kiela, Arthur Szlam, Jason Weston
2019 arXiv   pre-print
In particular, we show that grounding on the details of the local environment, including location descriptions, and the objects (and their affordances) and characters (and their previous actions) present  ...  We describe the results of training state-of-the-art generative and retrieval models in this setting.  ...  the training set, and the actual content (descriptions, personas, dialogues) are entirely disjoint.  ... 
arXiv:1903.03094v1 fatcat:cvbsurxe2fcbfimngylgt7e5di
« Previous Showing results 1 — 15 out of 83,998 results