Entrepôt de données IRD: un service en ligne pour l'ouverture et le partage des données scientifiques au Sud - Archive ouverte HAL Accéder directement au contenu
Poster De Conférence Année : 2018

Entrepôt de données IRD: un service en ligne pour l'ouverture et le partage des données scientifiques au Sud

Résumé

Researchers are facing increasingly complex challenges: climate change, health, migration. This is all the more true if we look at inter-tropical areas, which are high-stakes regions (population growth, food security, increased pressure on resources). These challenges require interdisciplinarity, openness and the sharing of data and skills. In this context, research methods are evolving to adapt or benefit by providing research environments that can link data from different sources, new approaches that exploit big data. In addition, scientists must meet increasing requirements for the provability and reproducibility of their research experiments. To this end, new forms of publications have emerged. They are associated with new editorial policies, as is the case with data papers. Finally, legislators have recently addressed these concerns. For example, a recent European directive requires that all data from Horizon 2020 projects be opened and managed so that they can be FAIR (Findable, Accessible, Interoperable, Reusable). In this context and as a research actor in southern countries, IRD has started a project for an open data repository. The objective of this project is to ensure an open, sustainable sharing of the datasets of interest collected in intertropical areas by IRD and its partners. More precisely, the short-term objectives of this repository are to better preserve and enhance the informational heritage of the IRD and its partners. It is also about providing an identification, repository and data discovery services for researchers so that they can control its diffusion and sharing according to FAIR principles. In the longer term, the data repository aims to ensure the discovery of all the research data produced by IRD and managed in other existing repositories, databases or data centres. The aim is to provide the most comprehensive overview possible of the data produced. This project complements the databases hosted and distributed within the observatory systems or data centres existing in the IRD's intervention areas. The targeted data are those called the long tail of data. They correspond to data that are currently not used to produce new interpretations or new knowledge. Currently, the focus (funding, research) is on big data and has put aside small datasets. In the context of open science and interdisciplinary research, the important value of small and medium-sized datasets is of great importance. These data sets must be taken into account in the implementation of national and international e-infrastructures. To this end, the IRD wishes to participate through this institutional project in the effort to remobilize these small data sets. After having recalled the diversity of disciplinary fields covered by the IRD, the importance of the information heritage accumulated over decades in intertropical zones, we will shed light on current data management practices in the research domain. Based on some examples from the work of the research teams (genomics, fisheries, population census), the different needs and constraints encountered at the IRD to which the data repository must respond will be presented. In a second step, we will present the Dataverse tool on which we want to build the data repository. Developed by Harvard University and used by many research institutes around the world, it provides a wide range of functionalities to ensure: the identification, description and publication of data sets, whatever the discipline. It also has ergonomic user interfaces that allow the discovery of data sets from facets specific to the disciplinary field. Built on the Datacite standard and implementing the OAI-PMH protocol, it allows to be harvested by academic search engines or more broadly ensures the visibility of data via the search engine such as google data search engine. To strengthen interoperability between the different repositories and databases, particular attention will be paid to the use of and access to existing terminology repositories. Finally, we will provide the organizational elements we wish to put in place to ensure the description and deposit of data by researchers. On this point, our main concerns will be to ensure the training and support of researchers in good practices for the management and diffusion of their data. We will use the FAIR principles to provide interoperable, well-documented data associated with appropriate user licenses. To this end, the support of the deposit of a data will be ensured as much on the relevance of the description of the data as on the choice of the data standard format or on the legal aspects relating to the opening of the data. We will conclude our communication by emphasizing the importance of such a project for the IRD, particularly through its research support missions in the south. We will discuss the perspectives of deploying such a framework (tool, data, management practices) to IRD partners in West Africa.
Les scientifiques font face à des défis de plus en plus complexes : changement climatique, santé, migrations. Ceci est d’autant plus vrai si l’on s’intéresse aux zones intertropicales qui sont des régions à forts enjeux (croissance démographique, sécurité alimentaire, pression accrue sur les ressources). Ces défis nécessitent l'interdisciplinarité, l’ouverture et la mutualisation des données et des compétences. Dans ce contexte, les méthodes de la recherche évoluent pour s’adapter ou tirer en profit en proposant des environnements de recherche capables de lier des données provenant de différentes sources, de nouvelles approches exploitant les données massives. Dans ce contexte et tant qu’acteur de recherche dans les pays du sud, l’IRD a démarré un projet d’entrepôt de données ouvertes. L’objectif de ce projet est d’assurer un partage ouvert, pérenne des jeux de données d’intérêt récoltées dans les zones intertropicales par l’IRD et ses partenaires. Plus précisément, les objectifs à court terme de cet entrepôt sont de mieux pérenniser et valoriser le patrimoine informationnel de l’IRD et de ses partenaires. Il s'agit également de fournir un service de d’identification, de dépôts et de découverte des données pour les scientifiques afin qu’ils puissent en maîtriser la diffusion et le partage selon les principes FAIR. A plus long terme, l’entrepôt de données vise à assurer la découverte des ressources numériques produites par l’IRD (données, support multimédia, publications) et déposées dans d’autres dépôts, bases de données, pôles de données. Il s’agit de fournir une vision la plus exhaustive possible du patrimoine de données.
Fichier principal
Vignette du fichier
poster-JSO2018-IRD-dataverse_HH.pdf (1.31 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-01951023 , version 1 (14-04-2023)

Identifiants

Citer

Jean-Christophe Desconnets, Pascal Aventurier, Stéphane Banon, Caroline Doucouré, Thibault Coupin, et al.. Entrepôt de données IRD: un service en ligne pour l'ouverture et le partage des données scientifiques au Sud. Premières journées de la science ouverte, Dec 2018, Paris, France. ⟨hal-01951023⟩
38 Consultations
16 Téléchargements

Partager

Gmail Facebook X LinkedIn More