Frédéric Béchet

Sebastien Montella, Alexis Nasr, Johannes Heinecke et al.

Text generation from Abstract Meaning Representation (AMR) has substantially benefited from the popularized Pretrained Language Models (PLMs). Myriad approaches have linearized the input graph as a sequence of tokens to fit the PLM tokenization requirements. Nevertheless, this transformation jeopardizes the structural integrity of the graph and is therefore detrimental to its resulting representation. To overcome this issue, Ribeiro et al. have recently proposed StructAdapt, a structure-aware adapter which injects the input graph connectivity within PLMs using Graph Neural Networks (GNNs). In this paper, we investigate the influence of Relative Position Embeddings (RPE) on AMR-to-Text, and, in parallel, we examine the robustness of StructAdapt. Through ablation studies, graph attack and link prediction, we reveal that RPE might be partially encoding input graphs. We suggest further research regarding the role of RPE will provide valuable insights for Graph-to-Text generation.

Ismael Rousseau, Geraldine Damnati, Frederic Bechet

While search agents demonstrate impressive capabilities in multi-step question answering, their robustness to poor-quality evidence remains under-explored. This phenomenon occurs rarely in realistic benchmarks but can lead to dramatic failure in real life applications. Therefore in this study we propose DeepStress, a stress testing framework that controls the frequency of challenging evidence by replacing the retrieval module of search agents with a controlled synthetic environment. We use this framework to control three dimensions that can affect document reliability: trustworthiness, relevance, and factuality. Testing several search agents on HotpotQA and BrowseCompPlus, we demonstrate that agents exhibit substantial differences in their ability to handle unreliable information and propose new metrics that better document systems outcomes as well as the interactions between conflicting parametric and retrieved knowledge.

Eunice Akani, Benoit Favre, Frederic Bechet et al.

Dialogue summarization aims to provide a concise and coherent summary of conversations between multiple speakers. While recent advancements in language models have enhanced this process, summarizing dialogues accurately and faithfully remains challenging due to the need to understand speaker interactions and capture relevant information. Indeed, abstractive models used for dialog summarization may generate summaries that contain inconsistencies. We suggest using the semantic information proposed for performing Spoken Language Understanding (SLU) in human-machine dialogue systems for goal-oriented human-human dialogues to obtain a more semantically faithful summary regarding the task. This study introduces three key contributions: First, we propose an exploration of how incorporating task-related information can enhance the summarization process, leading to more semantically accurate summaries. Then, we introduce a new evaluation criterion based on task semantics. Finally, we propose a new dataset version with increased annotated data standardized for research on task-oriented dialogue summarization. The study evaluates these methods using the DECODA corpus, a collection of French spoken dialogues from a call center. Results show that integrating models with task-related information improves summary accuracy, even with varying word error rates.

Hichem Ammar Khodja, Frédéric Béchet, Quentin Brabant et al.

The factuality of large language model (LLMs) tends to decay over time since events posterior to their training are "unknown" to them. One way to keep models up-to-date could be factual update: the task of inserting, replacing, or removing certain simple (atomic) facts within the model. To study this task, we present WikiFactDiff, a dataset that describes the evolution of factual knowledge between two dates as a collection of simple facts divided into three categories: new, obsolete, and static. We describe several update scenarios arising from various combinations of these three types of basic update. The facts are represented by subject-relation-object triples; indeed, WikiFactDiff was constructed by comparing the state of the Wikidata knowledge base at 4 January 2021 and 27 February 2023. Those fact are accompanied by verbalization templates and cloze tests that enable running update algorithms and their evaluation metrics. Contrary to other datasets, such as zsRE and CounterFact, WikiFactDiff constitutes a realistic update setting that involves various update scenarios, including replacements, archival, and new entity insertions. We also present an evaluation of existing update algorithms on WikiFactDiff.

Ikram Belmadani, Oumaima El Khettari, Carlos Ramisch et al.

The development of large language models (LLMs) has led to an increased focus on their adaptation to specialized domains and languages, yet the effectiveness of domain adaptation strategies remains unclear. We present a study of medical domain adaptation using French medical question-answering (QA) as a case study. We compare continual pretraining (CPT), supervised fine-tuning (SFT), and their combination across three model families, multiple sizes, and three initialization types, explicitly disentangling adaptation effects from base model choice. We evaluate both multiple-choice (MCQA) and open-ended QA (OEQA) under greedy and constrained decoding using automatic metrics and LLM-as-a-Judge evaluation. For MCQA, CPT+SFT most often achieves the best scores, but gains over SFT are small and frequently not statistically significant, making SFT a strong and cost-effective default. For OEQA, CPT consistently improves overlap-based metrics, while SFT often degrades generation quality; instruction tuning and CPT+SFT are preferred by LLM-based evaluation. Cross-lingual experiments further show effective transfer from French adaptation to English benchmarks. Overall, we provide practical guidelines for selecting adaptation strategies under computational constraints.

Elie Antoine, Frédéric Béchet, Philippe Langlais

This study investigates the behavior of model-integrated routers in Mixture of Experts (MoE) models, focusing on how tokens are routed based on their linguistic features, specifically Part-of-Speech (POS) tags. The goal is to explore across different MoE architectures whether experts specialize in processing tokens with similar linguistic traits. By analyzing token trajectories across experts and layers, we aim to uncover how MoE models handle linguistic information. Findings from six popular MoE models reveal expert specialization for specific POS categories, with routing paths showing high predictive accuracy for POS, highlighting the value of routing paths in characterizing tokens.

Hichem Ammar Khodja, Frédéric Béchet, Quentin Brabant et al.

This paper explores the robustness of language models (LMs) to variations in the temporal context within factual knowledge. It examines whether LMs can correctly associate a temporal context with a past fact valid over a defined period, by asking them to differentiate correct from incorrect contexts. The LMs' ability to distinguish is analyzed along two dimensions: the distance of the incorrect context from the validity period and the granularity of the context. To this end, a dataset called TimeStress is introduced, enabling the evaluation of 18 diverse LMs. Results reveal that the best LM achieves a perfect distinction for only 11% of the studied facts, with errors, certainly rare, but critical that humans would not make. This work highlights the limitations of current LMs in temporal representation.

Elie Antoine, Frédéric Béchet, Géraldine Damnati et al.

We introduce an evaluation methodology for reading comprehension tasks based on the intuition that certain examples, by the virtue of their linguistic complexity, consistently yield lower scores regardless of model size or architecture. We capitalize on semantic frame annotation for characterizing this complexity, and study seven complexity factors that may account for model's difficulty. We first deploy this methodology on a carefully annotated French reading comprehension benchmark showing that two of those complexity factors are indeed good predictors of models' failure, while others are less so. We further deploy our methodology on a well studied English benchmark by using Chat-GPT as a proxy for semantic annotation. Our study reveals that fine-grained linguisticallymotivated automatic evaluation of a reading comprehension task is not only possible, but helps understand models' abilities to handle specific linguistic characteristics of input examples. It also shows that current state-of-the-art models fail with some for those characteristics which suggests that adequately handling them requires more than merely increasing model size.

Gabriel Marzinotto, Geraldine Damnati, Frédéric Béchet

This paper presents a new semantic frame parsing model, based on Berkeley FrameNet, adapted to process spoken documents in order to perform information extraction from broadcast contents. Building upon previous work that had shown the effectiveness of adversarial learning for domain generalization in the context of semantic parsing of encyclopedic written documents, we propose to extend this approach to elocutionary style generalization. The underlying question throughout this study is whether adversarial learning can be used to combine data from different sources and train models on a higher level of abstraction in order to increase their robustness to lexical and stylistic variations as well as automatic speech recognition errors. The proposed strategy is evaluated on a French corpus of encyclopedic written documents and a smaller corpus of radio podcast transcriptions, both annotated with a FrameNet paradigm. We show that adversarial learning increases all models generalization capabilities both on manual and automatic speech transcription as well as on encyclopedic data.

Marc El-Bèze, Juan-Manuel Torres-Moreno, Frédéric Béchet

We present a set of probabilistic models applied to binary classification as defined in the DEFT'05 challenge. The challenge consisted a mixture of two differents problems in Natural Language Processing : identification of author (a sequence of François Mitterrand's sentences might have been inserted into a speech of Jacques Chirac) and thematic break detection (the subjects addressed by the two authors are supposed to be different). Markov chains, Bayes models and an adaptative process have been used to identify the paternity of these sequences. A probabilistic model of the internal coherence of speeches which has been employed to identify thematic breaks. Adding this model has shown to improve the quality results. A comparison with different approaches demostrates the superiority of a strategy that combines learning, coherence and adaptation. Applied to the DEFT'05 data test the results in terms of precision (0.890), recall (0.955) and Fscore (0.925) measure are very promising.

Gabriel Marzinotto, Frédéric Béchet, Géraldine Damnati et al.

This paper describes a Semantic Frame parsing System based on sequence labeling methods, precisely BiLSTM models with highway connections, for performing information extraction on a corpus of French encyclopedic history texts annotated according to the Berkeley FrameNet formalism. The approach proposed in this study relies on an integrated sequence labeling model which jointly optimizes frame identification and semantic role segmentation and identification. The purpose of this study is to analyze the task complexity, to highlight the factors that make Semantic Frame parsing a difficult task and to provide detailed evaluations of the performance on different types of frames and sentences.

Gabriel Marzinotto, Géraldine Damnati, Frederic Bechet

This article presents an automatic frame analysis system evaluated on a corpus of French encyclopedic history texts annotated according to the FrameNet formalism. The chosen approach relies on an integrated sequence labeling model which jointly optimizes frame identification and semantic role segmentation and identification. The purpose of this study is to analyze the task complexity from several dimensions. Hence we provide detailed evaluations from a feature selection point of view and from the data point of view.

Hussam Hamdan, Patrice Bellot, Frederic Bechet

Term weighting metrics assign weights to terms in order to discriminate the important terms from the less crucial ones. Due to this characteristic, these metrics have attracted growing attention in text classification and recently in sentiment analysis. Using the weights given by such metrics could lead to more accurate document representation which may improve the performance of the classification. While previous studies have focused on proposing or comparing different weighting metrics at two-classes document level sentiment analysis, this study propose to analyse the results given by each metric in order to find out the characteristics of good and bad weighting metrics. Therefore we present an empirical study of fifteen global supervised weighting metrics with four local weighting metrics adopted from information retrieval, we also give an analysis to understand the behavior of each metric by observing and analysing how each metric distributes the terms and deduce some characteristics which may distinguish the good and bad metrics. The evaluation has been done using Support Vector Machine on three different datasets: Twitter, restaurant and laptop reviews.