Focus on Lillodata (Université de Lille)
The University of Lille’s institutional data hub was established in the summer of 2022 and renamed Lillodata the following year.
The hub is managed by four members of the LORD (Lille Open Research Data) data cluster management. However, the scopes are not entirely identical. Indeed, within the scope of the LORD data management cluster, several institutional repositories coexist: those of IMT Nord Europe and UPHF, and those of the national research organisations (ONRs) active in the region (Inria, INRAE, CNRS).
It currently hosts 160 datasets comprising 2,034 files, as well as collections corresponding to 26 laboratories and 5 research projects.
Interview with Alicia León y Barella, head of the LORD data cluster management
Which disciplines are primarily represented by this data?
Just like the University of Lille, all disciplines are represented in Lillodata. However, there is a predominance of experimental or simulation data from the experimental sciences and engineering, which are likely less well covered by trusted thematic repositories. These disciplines account for roughly half of the datasets deposited in Lillodata. They relate more specifically to applied mathematics, materials physics, and engineering sciences (civil engineering, materials engineering – particularly textiles –, fluid mechanics, electronics and micro-nano-technologies).
What is the process for submitting data? For creating a collection?
Members of research units overseen by the University of Lille can log in to Lillodata and have the rights to submit a dataset.
To create collections, the four Lillodata administrators and a number of curators for spaces dedicated to laboratories are responsible for creating the collections. We create collections on request for research units, research teams, research projects, scientific journals, etc.
What curation is carried out on the platform?
We provide documentary and legal curation of datasets: adding metadata (notably author and affiliation), quality control of the submission, and checking the legal compliance of the submission (in the case of personal data, liaising with our DPO team where necessary). A curation report, adapted from that of Recherche Data Gouv, is systematically sent to the depositor upon completion of the deposit and prior to validation and publication online. We are sometimes under time pressure, as the deposit of a dataset may be linked to the publication of an article, in which case we must act swiftly so as not to delay the editorial process.
Is there any integration with other repositories?
As is the case everywhere, our units fall under the remit of several different bodies (other universities and the ONR). However, it is important for our unit directors to have access to a relatively comprehensive collection of datasets deposited on Recherche Data Gouv and produced by researchers affiliated with their unit (regardless of which body they are attached to). Consequently, after some trial and error and discussions with our regular partners (notably the teams managing the CNRS, UPHF and Inria portals), we have finally opted for an annual review, enabling us to link datasets affiliated with one of our units but deposited elsewhere, so as to make them visible on Lillodata and within the collections of our research units.
Do you have a model dataset or one you would recommend as an example of a ‘clean’ deposit?
From a documentary (and scientific) perspective, I find the work of the ALGOMUS team, affiliated with the CRIStAL laboratory, particularly interesting; they work at the intersection of musicology and computer science.
The datasets bring together digital corpora dedicated to the algorithmic analysis of music from the perspective of digital humanities and computational musicology. They are based on musical corpora enriched with analytical annotations, structured metadata and various digital objects (scores, recordings, synchronised analyses), which make them particularly original from a documentary perspective.
The data is closely linked to the Dezrann platform, an environment for visualising and annotating musical corpora: the data deposited on Lillodata is partly encoded in formats specific to this platform, enabling synchronisation between scores, audio recordings and analytical annotations, and thus facilitating their scientific use.
Several other datasets are currently being finalised and will be published on Lillodata: at that point, we plan to give them specific prominence on the Recherche Data Gouv portal (what a teaser!)
Discover the Algomus collection: https://entrepot.recherche.data.gouv.fr/dataverse/cristal-algomus
URL for the Institutional Portal contact desk: lillodata[at]univ-lille.fr
URL for the institutional section: https://entrepot.recherche.data.gouv.fr/dataverse/lillodata
Find out more about the LORD data workshop: https://lord.univ-lille.fr