Filters








31,311 Hits in 3.0 sec

Similarity Scoring for Dialogue Behaviour Comparison

Stefan Ultes, Wolfgang Maier
2020 SIGDIAL Conferences  
In this paper, we propose a general methodology to compute the similarity of two dialogue behaviour models and investigate different ways of computing scores on both the semantic and the textual level.  ...  For instance, two models may have a similar task success rate, but very different ways of getting there.  ...  Similarity Scores A similarity score is computed for the comparison of two behavioural models π and π .  ... 
dblp:conf/sigdial/UltesM20 fatcat:2oj3sveunfcarbbmygoav6tkka

A survey on metrics for the evaluation of user simulations

Olivier Pietquin, Helen Hastie
2012 Knowledge engineering review (Print)  
However, such data are generally required for designing, training and assessing dialogue systems.  ...  User simulations are especially needed when using machine learning methods for optimizing dialogue management strategies such as Reinforcement Learning, where the amount of data necessary for training  ...  The authors thank the referees, as well as Oliver Lemon and Paul Crook for their help and comments.  ... 
doi:10.1017/s0269888912000343 fatcat:ixyevgajl5fcnhnvsmx6vnchsy

Evaluation of a hierarchical reinforcement learning spoken dialogue system

Heriberto Cuayáhuitl, Steve Renals, Oliver Lemon, Hiroshi Shimodaira
2010 Computer Speech and Language  
a reasonable hand-coded strategy; and (c) hierarchical reinforcement learning dialogue agents are feasible and promising for the (semi) automatic design of optimized dialogue behaviours in larger-scale  ...  These dialogues were used to evaluate three types of machine dialogue behaviour: hand-coded, fully-learnt and semi-learnt.  ...  Eq. (20) is suited for simple slot-filling applications, but for more complex dialogues more rules have to be added.  ... 
doi:10.1016/j.csl.2009.07.001 fatcat:hrjxs2dv3ja4jioj4p56mz6h6q

A Sequence-to-Sequence Model for User Simulation in Spoken Dialogue Systems [article]

Layla El Asri and Jing He and Kaheer Suleman
2016 arXiv   pre-print
Previous models for user simulation suffer from several drawbacks, such as the inability to take dialogue history into account, the need of rigid structure to ensure coherent user behaviour, heavy dependence  ...  User simulation is essential for generating enough data to train a statistical spoken dialogue system.  ...  Notably, the F-score for these models is similar to the F-score of the agenda-based model on the test set of DSTC2.  ... 
arXiv:1607.00070v1 fatcat:s3jprgadpjds3cyeilc3jbx5se

A Sequence-to-Sequence Model for User Simulation in Spoken Dialogue Systems

Layla El Asri, Jing He, Kaheer Suleman
2016 Interspeech 2016  
Previous models for user simulation suffer from several drawbacks, such as the inability to take dialogue history into account, the need of rigid structure to ensure coherent user behaviour, heavy dependence  ...  User simulation is essential for generating enough data to train a statistical spoken dialogue system.  ...  Notably, the F-score for these models is similar to the F-score of the agenda-based model on the test set of DSTC2.  ... 
doi:10.21437/interspeech.2016-1175 dblp:conf/interspeech/AsriHS16 fatcat:khrdlnhm5zgp5kjwqerlwu7ba4

Computer mediated interaction in Asperger's syndrome: the Bubble Dialogue program

Gnanathusharan Rajendran, Peter Mitchell
2000 Computers & Education  
Additionally, the blind ratings revealed that only one of the'Asperger' Bubble Dialogue scripts was dierent from the scripts generated by individuals with emotional and behavioural diculties.  ...  scripts produced by individuals with Asperger's syndrome and scripts produced by individuals with emotional and behavioural diculties.  ...  The authors would also like to thank the students, at the School of Psychology, University of Birmingham, for their work in rating the Bubble Dialogue scripts.  ... 
doi:10.1016/s0360-1315(00)00031-2 fatcat:63qoavtlcbcbhjwgh7lz6nxmvu

Analysis of Empathic Dialogue in Actual Doctor-Patient Calls and Implications for Design of Embodied Conversational Agents

Sana Salman, Deborah Richards, Patrina Caldwell
2021 Italian Journal of Computational Linguistics  
Empathy revolves around a wide range of verbal and non-verbal behaviours that include, for example, the choice of words in social dialogues.  ...  for health practitioners and future ECA dialogue development.  ...  A similarity scale would determine the threshold, and based on the similarity scores further empathic cues could be recommended.  ... 
doi:10.4000/ijcol.862 fatcat:fh24sntyxzdu5dkalojqedixzy

Dialogue state tracking accuracy improvement by distinguishing slot-value pairs and dialogue behaviour

Khaldoon H. Alhussayni, Alexander Zamyatin, S. Eman Alshamery
2021 Indonesian Journal of Electrical Engineering and Computer Science  
The implementation was attempted with the prior user utterance as a dialogue encoder and second by the additional score combined with all the candidate slot-value pairs in the context of previous user  ...  for that are ineffective.  ...  Zhong et al. (2018) proposed the GLAD model with self-attentionbased recurrent networks and simulated representations for each uterus and previous system behaviour, calculating a similarity to each slot-value  ... 
doi:10.11591/ijeecs.v21.i2.pp1057-1064 fatcat:2swl6c6obvenbenqt56akn3qb4

Investigating Online Dialogue Journal Writing Impacts on Low Proficiency Students' Writing Anxiety

Anne Rowena David, Hazita Azman, Thang Siew Ming
2018 International Journal of Language Education and Applied Linguistics  
Second Language Writing Anxiety Inventory (SLWAI) was employed in the study to measure students' writing anxiety and also types of writing anxiety, namely somatic anxiety, cognitive anxiety and avoidance behaviour  ...  This case study required students to write dialogue journal entries to the researcher and to their partner based on topics provided in the university's e-learning portal (eLearn@USM).  ...  A similar trend was also mirrored in avoidance behaviour as the mean score was significant and there was also a change from high writing anxiety level to the moderate level.  ... 
doi:10.15282/ijleal.v8.1051 fatcat:syjcravydveu3evuv6uji6qjzi

Towards a General, Continuous Model of Turn-taking in Spoken Dialogue using LSTM Recurrent Neural Networks

Gabriel Skantze
2017 Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue  
The model is trained on human-human dialogue data to predict upcoming speech activity in a future time window.  ...  We show how this general model can be applied to two different tasks that it was not specifically trained for.  ...  Acknowledgements This work is supported by the SSF (Swedish Foundation for Strategic Research) project COIN, and the Swedish research council (VR) project Online learning of turn-taking behaviour in spoken  ... 
doi:10.18653/v1/w17-5527 dblp:conf/sigdial/Skantze17 fatcat:wsuszudwh5fu3b52lj5ux6zppu

Pattern discovery in annotated dialogues using dynamic programming

Thierry Lecroq, Alexandre Pauchet, Émilie Chanoni, Gerardo Ayala Solano
2012 International Journal of Intelligent Information and Database Systems  
This study describes a new heuristic to discover behavioural patterns in annotated dialogues: firstly, dialogues are transcribed and annotated; then, two-dimensional patterns are extracted; finally, the  ...  Recurrent patterns are extracted in the annotated dialogues using dynamic programming with the help of a substitution matrix specifically designed for the task.  ...  We obtain a series of pairs of alignments and the corresponding similarity score for each pair.  ... 
doi:10.1504/ijiids.2012.050097 fatcat:y5qy2moxljajffyd6abtgrno6q

Gaze, Dominance and Dialogue Role in the MULTISIMO Corpus

Lorcan McLaren, Maria Koutsombogera, Carl Vogel
2020 2020 11th IEEE International Conference on Cognitive Infocommunications (CogInfoCom)  
Using a multi-modal task oriented dialogue corpus, we evaluate relations between gaze behaviours, conversational dominance and dialogue role.  ...  When aligned with dialogue roles, gaze behaviour differences between players are demonstrated when occupying the role of speaker but not of listener.  ...  Using a multi-modal task oriented dialogue corpus, we evaluate relations between gaze behaviours, conversational dominance and dialogue role.  ... 
doi:10.1109/coginfocom50765.2020.9237833 fatcat:it65sadwv5dlhpyrbr7cjwsk4u

Adapting a Virtual Advisor's Verbal Conversation Based on Predicted User Preferences: A Study of Neutral, Empathic and Tailored Dialogue

Hedieh Ranjbartabar, Deborah Richards, Ayse Aysin Bilgin, Cat Kutay, Samuel Mascarenhas
2020 Multimodal Technologies and Interaction  
Analyses of rule accuracy, participants' dialogue preferences, and individual differences reveal that the three groups had different needs for empathic dialogue and highlight the importance and challenges  ...  ., reduced stress, increased rapport) of tailoring dialogue to the specific user (Adaptive group) with dialogues that are either empathic (Empathic group) or neutral (Neutral group).  ...  Acknowledgments: Thanks to Meredith Porte for technical assistance and to all participants involved in the studies. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/mti4030055 fatcat:v2uadsgndvbklo4xddn4apzwf4

SentEmojiBot: Empathising Conversations Generation with Emojis [article]

Akhilesh Ravi, Amit Yadav, Jainish Chauhan, Jatin Dholakia, Naman Jain, Mayank Singh
2021 arXiv   pre-print
The increasing use of dialogue agents makes it extremely desirable for them to understand and acknowledge the implied emotions to respond like humans with empathy.  ...  However, none of the AI systems utilizes emojis for empathetic conversation generation.  ...  We hope our pipeline and results will promote more research on using cross-modality data like emojis for improving empathetic behaviour of dialogue agents.  ... 
arXiv:2105.12399v1 fatcat:q2qgkjaemrebfno4reauav3l4q

Blending Task Success and User Satisfaction: Analysis of Learned Dialogue Behaviour with Multiple Rewards

Stefan Ultes, Wolfgang Maier
2021 SIGDIAL Conferences  
Recently, principal reward components for dialogue policy reinforcement learning use task success and user satisfaction independently and neither the resulting learned behaviour has been analysed nor a  ...  In this work, we employ both principal reward components jointly and propose a method to analyse the resulting behaviour through a structured way of probing the learned policy.  ...  Furthermore, an optimal blend was selected for a gated multiobjective reward function supported by similarity scores leading to a good balance between user satisfaction and task success.  ... 
dblp:conf/sigdial/UltesM21 fatcat:q2ijvpuenrhtrca7hnzsp7hj2a
« Previous Showing results 1 — 15 out of 31,311 results