Zobrazeno 1 - 10
of 170
pro vyhledávání: '"Égré, Paul"'
In this paper, we propose an experimental investigation of the problem of AI fairness in classification. We train an AI model and develop our own fairness package FairDream to detect inequalities and then to correct for them, using income prediction
Externí odkaz:
http://arxiv.org/abs/2407.14766
Autor:
Faye, Géraud, Casanova, Morgane, Icard, Benjamin, Chanson, Julien, Gadek, Guillaume, Gravier, Guillaume, Égré, Paul
This paper summarizes the experiments and results of the HYBRINFOX team for the CheckThat! 2024 - Task 1 competition. We propose an approach enriching Language Models such as RoBERTa with embeddings produced by triples (subject ; predicate ; object)
Externí odkaz:
http://arxiv.org/abs/2407.03850
Autor:
Casanova, Morgane, Chanson, Julien, Icard, Benjamin, Faye, Géraud, Gadek, Guillaume, Gravier, Guillaume, Égré, Paul
This paper presents the HYBRINFOX method used to solve Task 2 of Subjectivity detection of the CLEF 2024 CheckThat! competition. The specificity of the method is to use a hybrid system, combining a RoBERTa model, fine-tuned for subjectivity detection
Externí odkaz:
http://arxiv.org/abs/2407.03770
Autor:
Égré, Paul, Spector, Benjamin
Why is language vague? Vagueness may be explained and rationalized if it can be shown that vague language is more useful to speaker and hearer than precise language. In a well-known paper, Lipman proposes a game-theoretic account of vagueness in term
Externí odkaz:
http://arxiv.org/abs/2404.18154
Autor:
Icard, Benjamin, Maine, François, Casanova, Morgane, Faye, Géraud, Chanson, Julien, Gadek, Guillaume, Atemezing, Ghislain, Bancilhon, François, Égré, Paul
We present a corpus of 100 documents, OBSINFOX, selected from 17 sources of French press considered unreliable by expert agencies, annotated using 11 labels by 8 annotators. By collecting more labels than usual, by more annotators than is typically d
Externí odkaz:
http://arxiv.org/abs/2403.16099
Autor:
Faye, Géraud, Icard, Benjamin, Casanova, Morgane, Chanson, Julien, Maine, François, Bancilhon, François, Gadek, Guillaume, Gravier, Guillaume, Égré, Paul
This paper investigates the language of propaganda and its stylistic features. It presents the PPN dataset, standing for Propagandist Pseudo-News, a multisource, multilingual, multimodal dataset composed of news articles extracted from websites ident
Externí odkaz:
http://arxiv.org/abs/2402.03780
Autor:
Blomet, Quentin, Égré, Paul
The set of $\mathsf{ST}$-valid inferences is neither the intersection, nor the union of the sets of $\mathsf{K}_3$- and $\mathsf{LP}$-valid inferences, but despite the proximity to both systems, an extensional characterization of $\mathsf{ST}$ in ter
Externí odkaz:
http://arxiv.org/abs/2401.03436
Publikováno v:
The Review of Symbolic Logic 17 (2024) 682-704
Given a three-valued definition of validity, which choice of three-valued truth tables for the connectives can ensure that the resulting logic coincides exactly with classical logic? We give an answer to this question for the five monotonic consequen
Externí odkaz:
http://arxiv.org/abs/2312.16035
We present a hybrid approach to the automated measurement of vagueness and subjectivity in texts. We first introduce the expert system VAGO, we illustrate it on a small benchmark of fact vs. opinion sentences, and then test it on the larger French pr
Externí odkaz:
http://arxiv.org/abs/2309.06132
Autor:
Égré, Paul
This paper revisits Buridan's Bridge paradox (Sophismata, chapter 8, Sophism 17), itself close kin to the Liar paradox, a version of which also appears in Bradwardine's Insolubilia. Prompted by the occurrence of the paradox in Cervantes's Don Quixote
Externí odkaz:
http://arxiv.org/abs/2209.10625