Papers by Philippe Blache
Two-level classification for dialogue act recognition in task-oriented dialogues (2020.coling-main)
Copied to clipboard
| Challenge: | Existing methods for dialogue act classification are limited and feature sets are low . recognizing dialogue acts is useful for identifying type of information and knowledge to be conveyed . |
| Approach: | They propose a 2-level classification technique, distinguishing between generic and specific dialogue acts (DA) they propose an efficient approach for specific DA, based on high-level linguistic features. |
| Outcome: | The proposed method outperforms classical methods for DA classification by including high-level features. |
Comparing Probabilistic, Distributional and Transformer-Based Models on Logical Metonymy Interpretation (2020.aacl-main)
Copied to clipboard
| Challenge: | Logical metonymies are type clashes between an event-selecting verb and an entity-denoting noun . they are typically interpreted by inferring a hidden event on the basis of contextual cues . |
| Approach: | They propose to use probabilistic and distributional models to model logical metonymy interpretation . they compare models with the best Transformer-based models and some traditional distributional ones . |
| Outcome: | The proposed models perform well on a complex scenario, but low performance on some datasets suggests that logical metonymy is still a challenging phenomenon for computational modeling. |
Did You Get It? A Zero-Shot Approach to Locate Information Transfers in Conversations (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing models do not provide an efficient way to locate information that enters the common ground. |
| Approach: | They propose a method based on segmentation of a conversation into themes followed by their summarization and obtain the location of information transfers by computing the distance between the theme summary and the different utterances produced by a speaker. |
| Outcome: | The proposed method is based on the segmentation of a conversation into themes followed by their summarization and obtains the location of information transfers by computing the distance between the theme summary and the different utterances produced by a speaker. |
A Semi-autonomous System for Creating a Human-Machine Interaction Corpus in Virtual Reality: Application to the ACORFORMed System for Training Doctors to Break Bad News (L18-1)
Copied to clipboard
Magalie Ochs, Philippe Blache, Grégoire de Montcheuil, Jean-Marie Pergandi, Jorane Saubesty, Daniel Francon, Daniel Mestre
| Challenge: | Existing methods for training doctors to break bad news are expensive and time consuming. |
| Approach: | They propose a method to collect a corpus of human-machine interactions and then construct a semi-autonomous system based on the collected corpus. |
| Outcome: | The proposed system is based on a corpus-based method to analyze human-machine interactions and then develop fully autonomous prototype. |
Did the Cat Drink the Coffee? Challenging Transformers with Generalized Event Knowledge (2021.starsem-1)
Copied to clipboard
Paolo Pedinotti, Giulia Rambelli, Emmanuele Chersoni, Enrico Santus, Alessandro Lenci, Philippe Blache
| Challenge: | Prior work has explored the ability of computational models to predict word semantic fit with a given predicate. |
| Approach: | They compare Transformers Language Models to SDM to assess their performance . they found that TLMs do not capture important aspects of event knowledge . people can discriminate between typical and atypical events, they say . |
| Outcome: | The proposed models can achieve comparable performance to SDM, but they lack important aspects of event knowledge. |
The Badalona Corpus - An Audio, Video and Neuro-Physiological Conversational Dataset (2022.lrec-1)
Copied to clipboard
Philippe Blache, Salomé Antoine, Dorina De Jong, Lena-Marie Huttner, Emilia Kerr, Thierry Legou, Eliot Maës, Clément François
| Challenge: | Using the same dyads at different periods, we can study the evolution of interlocutors’ alignment during the time. |
| Approach: | They propose to record 5 dyads with all modalities and neuro-physiological signals in a natural conversation corpus. |
| Outcome: | The proposed corpus is the first to capture all modalities and neuro-physiological signals in a natural conversation situation. |
The Brain-IHM Dataset: a New Resource for Studying the Brain Basis of Human-Human and Human-Machine Conversations (2020.lrec-1)
Copied to clipboard
| Challenge: | Using a dataset of controlled interactions, we have studied the feedback items produced by the interlocutors during a conversation. |
| Approach: | They propose to use a dataset of controlled interactions to study feedback items and a virtual reality context to re-synthesize the conversations. |
| Outcome: | The proposed dataset compares human-human and human-machine production of feedbacks and is the first of its kind. |
The Distracted Ear: How Listeners Shape Conversational Dynamics (2024.lrec-main)
Copied to clipboard
| Challenge: | a new study examines the relationship between listener feedback and narration quality in human communication . a variety of complex linguistic and cognitive processes underpin the success of conversations . |
| Approach: | They analyze listener feedback, narration quality and distraction effects in a SMYLE corpus . they find a positive correlation between frequency of specific feedback and narration quality . |
| Outcome: | The proposed method shows that feedback plays a pivotal role in shaping the dynamics of conversations. |