Are cascade dialogue state tracking models speaking out of turn in spoken dialogues?

Autor: Druart, Lucas, Jacqmin, Léo, Favre, Benoît, Rojas-Barahona, Lina Maria, Vielzeuf, Valentin

In Task-Oriented Dialogue (TOD) systems, correctly updating the system's understanding of the user's needs is key to a smooth interaction. Traditionally TOD systems are composed of several modules that interact with one another. While each of these c

Externí odkaz: http://arxiv.org/abs/2311.04922

Active Learning and Multi-label Classification for Ellipsis and Coreference Detection in Conversational Question-Answering

Autor: Brabant, Quentin, Rojas-Barahona, Lina Maria, Gardent, Claire

In human conversations, ellipsis and coreference are commonly occurring linguistic phenomena. Although these phenomena are a mean of making human-machine conversations more fluent and natural, only few dialogue corpora contain explicit indications on

Externí odkaz: http://arxiv.org/abs/2207.03145

Bayesian Inverse Reinforcement Learning for Modeling Conversational Agents in a Virtual Environment

Autor: Rojas Barahona, Lina Maria, Cerisara, Christophe

Publikováno v: Lecture Notes on Computer Science
Conference on Intelligent Text Processing and Computational Linguistics
Conference on Intelligent Text Processing and Computational Linguistics, Alexander Gelbukh, Apr 2014, Kathmandu, Nepal

International audience; This work proposes a Bayesian approach to learn the behavior of hu- man characters that give advice and help users to complete tasks in a situated environment. We apply Bayesian Inverse Reinforcement Learning (BIRL) to in- fer

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::82e191893e5de46a43008566b60e6b13
https://hal.inria.fr/hal-01002361/document

Using Paraphrases and Lexical Semantics to Improve the Accuracy and the Robustness of Supervised Models in Situated Dialogue Systems

Autor: Gardent, Claire, Rojas Barahona, Lina Maria

Publikováno v: Conference on Empirical Methods in Natural Language Processing
Conference on Empirical Methods in Natural Language Processing, SIGDAT, the Association for Computational Linguistics special interest group on linguistic data and corpus-based approaches to NLP, Oct 2013, Seattle, United States. pp.808-813

International audience; This paper explores to what extent lemmatisation, lexical resources, distributional semantics and paraphrases can increase the accuracy of supervised models for dialogue management. The results suggest that each of these facto

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::45467c22bc1c1b7c11503d8182334f59
https://hal.inria.fr/hal-00905405

What should I do now? Supporting conversations in a serious game

Autor: Rojas Barahona, Lina Maria, Gardent, Claire

Publikováno v: SeineDial 2012-16th Workshop on the Semantics and Pragmatics of Dialogue
SeineDial 2012-16th Workshop on the Semantics and Pragmatics of Dialogue, Jonathan Ginzburg (chair), Anne Abeillé, Margot Colinet, Gregoire Winterstein, Sep 2012, Paris, France

International audience; We present two dialogue systems developed to support chating with French speaking virtual characters in the context of a serious game: one hybrid statistical/symbolic and one purely statistical. The player is guided in the que

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=dedup_wf_001::475e6406aa93af2627bec428c5597666
https://hal.inria.fr/hal-00757892/file/emospeech_demo.pdf