HAL Université de Toulouse, et Toulouse INP
Not a member yet
159308 research outputs found
Sort by
Confiance et fiabilité du cycle de vie décentralisé des données : une approche basée sur la blockchain pour améliorer la traçabilité des critères de qualité des données
No full textData science and AI-based techniques are now widely employed in various sectors, including business, politics, healthcare, transportation, research, etc. Private companies and public organizations have produced and/or collected various types of data which today are stored in data silos that need to be integrated to build a data economy that drives innovation such as data spaces in Europe. Such data spaces should involve different stakeholders in collaborative and distributed data processing, as well as decentralized data governance. However, these distributed and decentralized data ecosystems open up new challenges in terms of trust and reliability. Data quality evaluation is a potential indicator to assess the reliability, trust and legal compliance of shared data across collaborative data processing. Unfortunately, there are multiple and inconsistent definitions of data quality criteria, and data quality criteria are often hardly accessible to end-users. As an answer to these issues, we developed a framework that includes a comprehensive list of 30 data quality criteria obtained from an extensive systematic review of existing academic research articles and consolidated by the interviews of 9 experts. We manually assessed 4 different open source datasets according to 4 data quality criteria to demonstrate the usefulness of our approach as well as the difficulty to access this information currently. Therefore, we developed a blockchain-based traceability prototype that can improve transparency which allows any dataset user to easily access data quality criteria.Les techniques basées sur la science des données et l'IA sont désormais largement utilisées dans divers secteurs, notamment les affaires, la politique, la santé, les transports, la recherche, etc. Les entreprises privées et les organismes publics ont produit et/ou collecté différents types de données qui sont aujourd'hui stockées dans des silos de données et qui doivent être intégrés afin de créer une économie des données favorisant l'innovation, à l'image des espaces de données en Europe. Ces espaces de données devraient impliquer différentes parties prenantes dans un traitement collaboratif et distribué des données, ainsi qu'au travers d'une gouvernance décentralisée des données. Cependant, ces écosystèmes de données distribués et décentralisés posent de nouveaux défis en termes de confiance et de fiabilité. L'évaluation de la qualité des données est un indicateur potentiel permettant d'évaluer la fiabilité, la confiance et la conformité juridique des données partagées dans le cadre du traitement collaboratif des données. Malheureusement, il existe de multiples définitions incohérentes des critères de qualité des données, et ces derniers sont souvent difficilement accessibles aux utilisateurs finaux. Pour répondre à ces problèmes, nous avons développé un cadre comprenant une liste de 30 critères de qualité des données, obtenus à partir d'une revue systématique approfondie des articles de recherche universitaires existants et consolidés par les entretiens de 9 experts. Nous avons évalué manuellement 4 ensembles de données open source différents selon 4 critères de qualité des données afin de démontrer l'utilité de notre approche ainsi que la difficulté d'accéder à ces informations à l'heure actuelle. Nous avons enfin développé un prototype de traçabilité basé sur la blockchain qui peut améliorer la transparence et permettre à tout utilisateur d'ensembles de données d'accéder facilement à leurs critères de qualité des données
Atypical lymphoproliferations associated with germline genetic variants: a report of the 2024 EA4HP/SH lymphoma workshop
No full textInternational audienceSession 2 of the 2024 European Association for Haematopathology/Society for Hematopathology lymphoma workshop was dedicated to atypical lymphoproliferations in association with germline genetic variants. The first group of cases were lymphoproliferations occurring in the context of primary immunodeficiencies (PID), a heterogeneous group of diseases with increasing incidence and number of different diseases due to better recognition. The workshop contained a spectrum of different PIDs and associated lymphoproliferations with autoimmune lymphoproliferative syndrome, activated phosphoinositide 3-kinase delta syndrome, ataxia-telangiectasia and common variable immune deficiency being the most common. Both children and adults were affected, and the diagnosis of an underlying PID often required a high index of suspicion and correlation with clinical presentation and immunological/ infectious workup. Recognition of a PID allows specific treatment and can influence the interpretation of lymphoproliferations occurring in this context. The spectrum of lymphoproliferations ranged from reactive to overt lymphoma, both EBV-positive and -negative. In a subset of cases, it was very difficult or impossible to establish the boundary between reactive and neoplastic in the context of a PID. The second group represented a heterogeneous group of lymphoproliferations in the context of mutations in germline haematopoietic malignancy risk genes, without associated immunodeficiency. It was often difficult to determine if the genetic defect and the lymphoproliferation were causally related or coincidental, especially if the patient was also treated for non-lymphoid conditions. This is a rapidly evolving field in which future studies are expected to shed more light on the relationship between germline mutations and lymphoid malignancy.</div
Le projet « femmes géographes » sur Wikipédia : analyse d'une participation à un commun du savoir
No full textInternational audienc
Résolution d'entités pour les flux de données à l'aide de la technique d'embedding
No full textInternational audienceLa plupart des systèmes de traitement de flux de données recueillent des données provenant de différentes sources en temps réel et les consomment immédiatement. Cependant, de nombreuses analyses décisionnelles nécessitent des données en temps réel et des données historiques (par exemple, un dataset local ou des enregistrements antérieurs) en même temps pour comprendre la situation actuelle et avoir une vue globale. La résolution d'entités permet de déterminer si deux enregistrements différents font référence à la même entité en l'absence d'un identifiant unifié dans le cas multi-sources. Il est donc essentiel d'intégrer les données en temps réel aux données stockées en appliquant la résolution d'entités et de garantir l'accessibilité et la facilité d'utilisation des données multi-sources. Les méthodes existantes pour la résolution d'entités sont souvent incapables de prendre en charge ces dernières de manière continue tout en fournissant des méthodes générales et efficaces pour l'analyse de données complexes telles que le texte. Étant donné la bonne capacité de l'embedding dans la capture des informations sémantiques et syntaxiques, nous visons à appliquer cette technique au traitement de flux de données dans le but d'intégrer les données entrants en temps réel aux données historiques. En outre, la plupart des approches privilégient aujourd'hui la précision et la scalabilité tout en ignorant l'augmentation de la consommation d'énergie que nécessite l'amélioration de cette précision. Nous proposons donc ici une approche d'embedding de graphe dynamique adaptée au traitement des données en temps réel pour effectuer la résolution d'entités dans des tables relationnelles tout en évaluant la consommation d'énergie de ce traitement.</div
Evaluating Embeddable Language Models in Verbalizing Rule-based Inferences through Justifications
No full textInternational audienceWhile Language Models have shown promising performance, they still struggle with limitations regarding reasoning and are very token-sensitive. In contrast, knowledgebased systems, such as ontologies, allow for provable logically valid reasoning and provide explicit justifications regarding newly inferred knowledge. However, those justifications can be hard to understand for non-expert users given their formal syntax and their length. We investigated if language models could be considered as reliable tools for verbalizing such explanations, thus increasing explainability over reasoning output. This paper presents a reference evaluation of a set of embeddable language models on a task of translation from ontology formatted inferences and justifications into natural language sentences. We show that the order of justifications significantly decreases performance, whereas adding the inference rule as additional context significantly improves performance, leading to more reliable results
Mountain tourism stakeholders facing the climate, energy and decarbonisation challenges, perspectives and adaptations in the French Pyrenees
No full textInternational audienceClimate change is causing profound changes in mountain areas, with implications for tourism. The ski tourism sector has to adapt to these changes, face energy crises and the challenge of decarbonisation to meet the Paris Agreement targets. The French Pyrenees offers a relevant context for analysing the adaptation of ski tourism to environmental and energy challenges. In this context, the aim of this proposal is to explore how tourism stakeholders in the French Pyrenees perceive the climate and energy challenges and implement adaptation pathways. The research is interdisciplinary, based on management science and geography, and is based on 40 semi-structured interviews with institutional and private stakeholders from ski resorts in the Ariège and Haute Pyrenees. Preliminary findings show that most stakeholders agree that climate change is a challenge to their activities and recognise the need to adapt. However, most adaptation strategies remain reactive, and decarbonisation is often seen as a global objective. This point is reinforced by the fact that most of the carbon emissions from the ski resorts are due to the mobility of tourists. Stakeholders recognise the need to work on it, but are unable to act individually due to their areas of competence. The results confirm the need to strengthen the coordination between the different stakeholders and to involve the different territorial scales (from local to national)
A Fine-Grained Predictive Stress Quantification Framework for Drivers with Autism Spectrum Disorder in Inclusive Smart Mobility
No full textInternational audienceAutism Spectrum Disorder (ASD) is a neurodevelopmental condition that significantly heightens the challenges and risks associated with driving. Individuals with ASD often experience increased sensory sensitivities, difficulty adapting to dynamic environments, and heightened vulnerability to unexpected events, which can lead to elevated stress levels, sensory overload, and impaired decision-making on the road. Existing research in stress detection primarily proposes reactive approaches, triggering responses only after stress has escalated. Such strategies limit the possibility for timely and effective intervention. Moreover, most existing studies fail to quantify stress intensity, providing only binary assessments. This lack of granularity hinders the development of proactive and personalized support mechanisms. In this paper, we propose a novel real-time approach for continuous stress prediction and quantification tailored to drivers with ASD. Our method leverages physiological signals and contextual data to anticipate stress escalation and enable proactive stress quantification and provide a proactive, fine-grained assessment of stress levels, paving the way for personalized and context-aware in-vehicle support systems that enhance driving safety and comfort for individuals with ASD
Micro-indentation assessment of the impact of temperature and relative humidity conditions on cement paste visco-elastic properties
No full textInternational audienc
The Influence of Microstructural Heterogeneities on the Thermal Response of CFRTP Composite Tapes at the Ply-Scale
No full textInternational audienceThe thermal response of Carbon Fiber Reinforced Thermoplastic (CFRTP) tapes under short-term localized heating is critical for automated manufacturing processes. Conventional homogenized models often overlook microstructural heterogeneities that can promote non-uniform heating and affect the quality of the consolidated part. In this work, we combine insights from infrared thermography with finite element simulations at the fiber scale built on micrographs extracted from real tapes to quantify the effect of individual heterogeneities—including surface roughness, thickness variation, fiber agglomeration, and porosity—on thermal propagation. Three modeling configurations were compared under identical conditions: a full microstructure model; a simplified geometry-aware model (where the real geometry is taken into the account, including the surface roughness and thickness variability, but the properties of the domain are considered as a homogeneous-equivalent material); and a homogeneous-equivalent baseline with flat borders and uniform thickness. Results show that porosity effects depend strongly on location and orientation: large, horizontally aligned pores near the heated surface produce the highest gradients. Surface roughness, on the other hand, exerts dominant effects on surface temperature non-uniformity with respect to thickness variation and fiber distribution. These findings demonstrate that accounting for microscale heterogeneities is essential to achieve more accurate, optimized, and application-tailored analyses of CFRTP tapes in advanced manufacturing
Disruption of putrescine export in experimentally evolved <i>Ralstonia pseudosolanacearum</i> enhances symbiosis with <i>Mimosa pudica</i>
No full textInternational audiencePolyamines are essential molecules across all domains of life, but their role as signaling molecules in host-microbe interactions is increasingly recognized. However, because they are produced by both the host and the microbe, their dual origin makes their functional dissection challenging. The plant pathogenRalstonia pseudosolanacearumGMI1000 secretes large amounts of putrescine bothin vitroand in the xylem sap of host plants. In this study, we investigated the genetic changes underlying its experimental evolution into a legume symbiont. We showed that thepaeAgene (RSc2277), which was repeatedly mutated during this process, encodes a putrescine exporter. Mutations inpaeAcompletely abolished putrescine excretionin vitroand enhanced bacterial proliferation within nodules during interaction with the legumeMimosa pudica. When these mutations occurred in symbionts already capable of intracellular infection, it further increased bacterial load in nodules and allowed the detection of nitrogenase activity. In addition,paeA-mutated symbionts modulated host gene expression towards a more functional symbiotic state by repressing defense-related genes and inducing nodule development genes. These nodule development genes include genes encoding leghemoglobins and an arginine decarboxylase, a key enzyme in plant putrescine biosynthesis. These results indicate that bacterial and plant putrescine have distinct functions in legume symbiosis and highlight the complex role of polyamines in plant-microbe interactions. Importance Rhizobia, the nitrogen-fixing symbionts of legumes, emerged through repeated and independent horizontal transfers of some essential symbiotic genes. However, these transfers alone are often insufficient to convert the recipient bacterium into a functional legume symbiont. In a laboratory experiment, we evolved the plant pathogenRalstonia pseudosolanacearuminto a nodulating and intracellularly infecting symbiont ofMimosa pudica. This transition required genomic modifications in the recipient bacterium to activate its acquired symbiotic potential. Here, we demonstrated that one of these key adaptive modifications is the inactivation of bacterial putrescine export. This polyamine, when produced by the microsymbiont, appears to act as a negative signal for the plant. This study provides new insights into the distinct roles of bacterial- and plant-derived putrescine in plant-microbe interactions, highlighting their functional divergence despite being produced by both organisms.</p