Portail HAL ENC (École nationale des chartes-PSL)
Not a member yet
3792 research outputs found
Sort by
Le multilatéralisme entre 1929 et 1939 : une perspective du Parlement français
No full textMultilateralism, as a central concept in international relations, was institutionalized only after the Second World War with the creation of the United Nations, yet its ideas and practices were already discernible during the interwar period. France, as a major power of the time, offers a pertinent case study: between 1929 and 1939, parliamentary debates reveal both the dynamics of domestic political negotiation and the ways in which multilateralism was expressed within Parliament. This research draws on French parliamentary archives in connection with the economic crisis, disarmament conferences, and the League of Nations. Methodologically, it employs artificial intelligence tools, in particular Retrieval-Augmented Generation, as well as text-processing techniques, to analyse a large corpus both quantitatively and qualitatively. The study thus seeks to better understand the manifestation of multilateralism within the French Parliament and to demonstrate the potential of digital methods in historical research.Le multilatéralisme, en tant que concept central des relations internationales, n’a été institutionnalisé qu’après la Seconde Guerre mondiale avec la création de l’ONU, mais ses idées et pratiques étaient déjà perceptibles durant l’entre-deux-guerres. La France, puissance majeure de l’époque, offre un cas d’étude pertinent : entre 1929 et 1939, les débats parlementaires révèlent à la fois la dynamique de la négociation politique interne et la manière dont le multilatéralisme s’exprimait au sein du Parlement. Cette recherche s’appuie sur les archives parlementaires françaises, en lien avec la crise économique, les conférences de désarmement et la Société des Nations. Sur le plan méthodologique, elle mobilise des outils d’intelligence artificielle, en particulier la Génération augmentée par récupération, ainsi que des techniques de traitement du texte, afin d’analyser un corpus volumineux à la fois quantitativement et qualitativement. L’étude vise ainsi à mieux comprendre la manifestation du multilatéralisme au sein du Parlement français et à montrer le potentiel des méthodes numériques dans la recherche historique
Building a Data Infrastructure for Social Science and Humanities: A double perspective on quality and community from Italy and France
No full textInternational audienceThe Social Sciences and Humanities (SSH) disciplinary sector encompasses research studies that share an epistemological commitment to the critical investigation of human experience, cultural expression, and social organization. SSH research contributes to the development of theoretical frameworks and methodological approaches that are essential for understanding complex societal transformations.European institutions, including the European Commission, recognized the importance of SSH researchers and research projects by integrating it into broader research frameworks, such as Horizon Europe and NextGeneration EU, and supporting international cooperation through initiatives like the European Research Infrastructure Consortium (ERIC). Yet, despite this support, SSH research faces distinct challenges when compared to STEM disciplines: i) sparsity and fragmentation of its data; ii) heterogeneity of sources - textual, audiovisual, or contextual - generated under diverse, often non-standard conditions; iii) isolated datasets within institutional silos, limiting their discoverability, accessibility and reuse. Such variability underscores a systemic challenge in ensuring data quality across SSH research, which involves evaluating the accuracy, completeness, consistency, and documentation of data to enable its meaningful interpretation, reuse, and long-term preservation.Moving towards data-intensive and AI empowered research, RIs will see a further shift of their role from data suppliers to primary users of machine enabled workflows. To ensure sustainability of workflows, given by their reproducibility and replicability, data quality becomes a fundamental aspect of good research and reliable results, in particular in the case of AI mediated processes and workflows.DARIAH.it, as part of the H2IOSC project, and DARIAH-FR, particularly through the IR* Huma-Num, are two national research infrastructures that exist to serve SSH researchers and research projects in their respective countries. Both initiatives are inscribed in the ERIC DARIAH-EU, which seeks to structure the development and interaction of national research infrastructures in Europe. Though both Italy and France have different approaches, the common problem remains: how to improve quality of data, metadata, paradata and related tools based on an approach involving users?ItalyIn Italy, DARIAH.it is working to strengthen the national infrastructure, focusing on promoting standards for data, services and workflows, as well as developing platforms to support users in producing and managing FAIR data and resources, access a wide range of services and combine them into meaningful scientific workflows requiring the interaction of different & independent services, also leveraging on AI modules (i.e. AI-mediated DH). Recently, due to the increasing number of attacks brought to different institutions (British Libraries) and resources (e.g.: Archive.org) dealing with cultural content - Italy started a process of transition towards Critical Infrastructures for SSH, bringing cybersecurity and resilience into DARIAH.it national infrastructure development plans.Being a socio-technological environment, and aiming to bring measurable advancements for the research community, the upgrade of the technical infrastructures requires a strong investment on the human component: RIs can support research by providing quality data, tools and processes but researchers are encouraged to be the owner of their own algorithms and to know how they work to ensure they get the right answers to the right questions, hence transparence, explainability, standardization, alignment and training are the drivers to successfully complete this transition.FranceHuma-Num, the French national infrastructure for SSH has built a robust technical research infrastructure fed by community feedback. After its first decade, certain aspects need to be updated, most notably, the approach to data & metadata quality in its research data repository, NAKALA. Considering the large amount of existing data in NAKALA (over 1.4 million files), it was deemed unfeasible to manually curate all existing and future deposits.. To implement this quality plan, and in addition to purely technical controls, Huma-Num relies on communities to develop a network of curators, in articulation with the national ecosystem Recherche Data Gouv,, with their harmony assured by the development of a curation guide. Alongside this, Huma-Num has extensively developed its documentation, with a focus on data preparation, and has organized a series of webinar sessions for users. Finally, Huma-Num launched a global content analysis of the repository to gain a better understanding of current practices and develop quality indicators. By combining technical developments with community feedback to improve quality, Huma-Num gradually evolves from a purely technical infrastructure to a knowledge infrastructure. Conclusion This paper will develop the two Italian and French approaches to building an infrastructure for SSH centered on quality and sustainability. The problems to be solved are quite similar despite differences in national organizations, but the crucial common point is the necessary involvement of users, without which actions are doomed to failure. These dual approaches are fruitful examples for others that are confronted with the thorny problem ensuring research data and metadata quality in SSH.ReferencesEdmond, Jennifer, ed. 2020. “Digital Technology and the Practices of Humanities Research.” Open Book Publishers. https://doi.org/10.11647/obp.0192.Lacagnina, Carlo, Romain David, Anastasija Nikiforova, Mari Elisa Kuusniemi, Cinzia Cappiello, Oliver Biehlmaier, Louise Wright, et al. «TOWARDS A DATA QUALITY FRAMEWORK FOR EOSC». Zenodo, 9 January 2023. https://doi.org/10.5281/zenodo.7515816.Bellini, Emanuele and Emiliano Degl’Innocenti. 2024. Transitioning SSH European Research Infrastructures to Critical Infrastructure Through Resilience. IEEE International Conference on Cyber Security and Resilience (CSR): pp. 801-806, doi: 10.1109/CSR61664.2024.10679383. Spadi, Alessia, Emiliano Degl’Innocenti, e Carmen Di Meo. 2024. «DARIAH.It: Data Integration Strategies and Solutions for Digital Resources Management and Research in the Arts and Humanities». Mimesis Journal 13 (2):119-34. https://doi.org/10.13135/2389-6086/9920.Edward J. Gray, Nicolas Larrousse. Huma-Num IR*, 10 Years of Building a Research Infrastructure at the European level. Huma-Num, 10 Years of Building a Research Infrastructure at the European level, 2024. ⟨halshs-04573643
Comme les pièces d’un puzzle : décomposer et recomposer les exempla dans l’œuvre d’Hélinand de Froidmont
No full textInternational audienc
Édition d'une lettre chiffrée de Charles Quint à Jean de Saint-Mauris (1547)
No full textÉdition de sourcesConservée de manière isolée à la Bibliothèque municipale de Nancy, une lettre chiffrée adressée en février 1547 par Charles Quint à son ambassadeur auprès du roi de France, Jean de Saint-Mauris, a fait l'objet en 2022 d'une cryptanalyse. À la suite de celle-ci, un travail de recherche (autres exemplaires de la lettre, autres lettres chiffrées) et d'édition a été mené afin de replacer cette lettre dans l'environnement épistolaire et cryptographique large qui était le sien. Cette édition permet ainsi de disposer, outre d'un accès complet au texte de la lettre, d'un aperçu des pratiques cryptographiques impériales du milieu du XVIe siècle
A Riddle in a Haystack: LLM Detection of Intricate Wordplays in Colette and Willy's Novels for Authorship Attribution
No full textInternational audienceThis study explores the integration of wordplay detection into authorship attribution techniques, applied to French novels, with a focus on the celebrated writer Colette. Her first novel, Claudine à l'École, was originally published under her husband's name, Willy, and has since been the subject of authorship controversy. Since no works can be definitively attributed to Willy alone, traditional attribution methods are limited. However, historical sources consistently highlight his fondness for intricate and erudite wordplay.We first use transformer-based emotion detection to filter out emotionally negative passages, which are less likely to contain wordplay. We then use large language model (LLM) annotations to identify potential wordplay instances. Testing our pipeline on Claudine à l'École revealed a small number of sophisticated wordplays, concentrated in sections previously linked to Willy's influence through linguistic profiling. These findings lend tentative support to the hypothesis that Colette's direct writing was only minimally affected by Willy and offer a novel approach to attributing authorship in collaborative literary settings, particularly in the context of Willy's ghostwriting "workshop" and its many outputs.Cette étude explore l’intégration de la détection de jeux de mots aux techniques d’attribution d’autorité, en se concentrant sur le cas de la célèbre écrivaine française Colette. Son premier roman, Claudine à l’École, a d’abord été publié sous le nom de son mari, Willy, et fait depuis l’objet d’une controverse sur sa véritable paternité. Aucun ouvrage ne pouvant être attribué de manière certaine à Willy seul, les méthodes traditionnelles d’attribution se trouvent limitées. Cependant, les sources historiques soulignent de manière récurrente son goût pour les jeux de mots érudits et complexes.Nous utilisons d’abord une détection des émotions basée sur des modèles de type transformer pour filtrer les passages émotionnellement négatifs, moins susceptibles de contenir des jeux de mots. Nous mobilisons ensuite les annotations générées par un grand modèle de langage (LLM) pour identifier les occurrences potentielles de jeux de mots. Testée sur Claudine à l’École, notre méthode a permis de repérer un petit nombre de jeux de mots sophistiqués, concentrés dans des passages précédemment associés à l’influence de Willy via des analyses de style linguistique. Ces résultats apportent un soutien prudent à l’hypothèse selon laquelle l’écriture propre de Colette aurait été peu affectée par Willy, et proposent une approche nouvelle de l’attribution d’auteur dans des contextes littéraires collaboratifs, notamment dans le cadre de l’« atelier » d’écriture fantôme de Willy et de ses nombreuses productions
FAIRly publishing your textual data with DoTS
No full textInternational audienceThis poster aims to present DoTS, a comprehensive and functional suite of tools for publishing corpora in compliance with the DTS specification, integrating backend, API responses, and frontend for the creation of adaptable websites
La société philarmonique de Munster : histoire sociale d'un ensemble musical alsacien au temps de l'industrialisation (fin du XVIIIe siècle-1914)
No full textIn 1859, under the patronage of the Hartmann industrial dynasty, a unique musical ensemble was born in the Haut-Rhin: the Société philharmonique of Munster. Bringing together a men's choir, a women's choir and an orchestra, this association welcomes musicians from diverse social backgrounds, making it a tool for social cohesion. However, this cohesion is challenged by the predominant role played by local notables and employers, raising an essential question: to what extent does this musical ensemble reflect and reinforce the social and gender inequalities that mark the industrial era? In the early 19th century, musical life in Munster was dominated by the Hartmann family. The industrialists thus accompanied the birth of the Philharmonic and supervised it with authority, alongside the musical director. While the concerts, parties and excursions organized by the Société philharmonique were moments of sharing and entertainment, they were nonetheless marked by inequality. Although the association survived until 1934, the First World War marked the end of the Hartmanns' golden age and their hold on local cultural life.En 1859, sous le patronage de la dynastie industrielle des Hartmann, naît dans le Haut-Rhin un ensemble musical singulier : la Société philharmonique de Munster. Réunissant un chœur d’hommes, un chœur de femmes et un orchestre, cette association accueille des musiciennes et musiciens issus d’horizons sociaux divers, ce qui fait d’elle un outil de cohésion sociale. Cette cohésion est cependant mise à l’épreuve par la place prépondérante qu’y occupent les notables et le patronat local, soulevant ainsi une question essentielle : dans quelle mesure cet ensemble musical reflète-t-il et renforce-t-il les inégalités sociales et de genre qui marquent l’époque industrielle ? Au début du XIXe siècle, la vie musicale de Munster est dominée par les Hartmann. Les industriels accompagnent ainsi la naissance de la Philharmonie et la supervisent avec autorité, aux côtés du directeur musical. Si les concerts, fêtes et excursions organisés par la Société philharmonique sont des moments de partage et de divertissement, ils restent néanmoins empreints d’inégalités. Bien que l’association survive jusqu’en 1934, la Première Guerre mondiale marque la fin de l’âge d’or des Hartmann et de leur emprise sur la vie culturelle locale
0-148 ou 40 années de Culture et recherche.: Esquisse et perspectives
No full textInternational audienceBased on an analysis of the abstracts and physical presentation of the journal Culture et recherche, this article aims to trace the broad outlines of its evolution since its creation in 1984 as a simple newsletter on cultural research, until its consolidation in the 2010s as a vehicle for disseminating the research policies promoted by the Ministry of Culture.The article is followed by a reproduction of an article previously published in the journal, which illustrates its conclusion: Hélène Hatzfeld and Sylvie Grange, “Questioning cultural policies in the face of the complexity of contemporary societies,” Culture et recherche, no. 128 (Interculturalism in action)Sur la base d'une analyse des résumés et de la présentation physique de la revue Culture et recherche, cet article vise à retracer les grandes lignes de son évolution depuis sa création en 1984 en tant que simple bulletin d'information sur la recherche culturelle, jusqu'à sa consolidation dans les années 2010 en tant que vecteur de diffusion des politiques de recherche promues par le ministère de la Culture.L'article est suivi de la reproduction d'un article précédemment publié dans la revue, qui illustre sa conclusion : Hélène Hatzfeld et Sylvie Grange, « Questionner les politiques culturelles face à la complexité des sociétés contemporaines », Culture et recherche, n° 128 (L'interculturel en actes)