Services & FAIR Data

Tools and services provided by Huma-Num

The Documentation

All Huma-Num services are thoroughly documented on the official documentation portal:

Tip📖 documentation.huma-num.fr

🔗 https://documentation.huma-num.fr

This is the authoritative reference for every service described on this page — tutorials, usage conditions, quotas, and contact information. Always start here when you encounter a problem or want to go further with any service.

Access to most web-based services (ShareDocs, GitLab, Mattermost, NAKALA, Isidore, Kanboard…) is managed via HumanID, the centralised authentication service: https://humanid.huma-num.fr

FAIR Principles in Practice

The FAIR principles — Findable, Accessible, Interoperable, Reusable — are central to Huma-Num’s mission and to open science policy in France and Europe.

Note

Since 2021, the Plan national pour la science ouverte requires that research data from publicly-funded projects be deposited in open, FAIR-compliant repositories. Horizon Europe grants impose similar requirements at the European level.

🔗 go-fair.org/fair-principles · DMP OPIDoR

Core Services

NAKALA — Research Data Repository

NAKALA is Huma-Num’s main data publishing platform. It allows researchers to deposit and share humanities research data openly, with:

  • Persistent ARK identifiers for each deposited item
  • Rich metadata description (Dublin Core, custom schemas)
  • Open access licences (Creative Commons)
  • FAIR-compliant infrastructure, connected to EOSC
  • NAKALA-Press — generate a public-facing website with a custom nom.nakala.fr domain directly from a NAKALA collection

🔗 https://nakala.fr · Documentation

COCOON — Oral Corpus Repository

COCOON (COllections de COrpus Oraux Numériques) is a specialised repository for oral and multimodal corpora, hosted by Huma-Num and managed in connection with the CORLI consortium. Unlike NAKALA, which is a general-purpose research data repository, COCOON is designed specifically for annotated speech data — interviews, conversations, fieldwork recordings — with metadata schemas adapted to oral material (speakers, languages, recording situations) and tools for managing the ethical and legal constraints inherent to this type of data (anonymisation, restricted access, consent). COCOON is integrated into the CLARIN infrastructure through CORLI’s Centre-K node.

🔗 https://cocoon.huma-num.fr

Isidore — Discovery Platform

Isidore is a search and discovery engine aggregating open access scholarly publications, data, and resources in the humanities and social sciences. It harvests metadata from hundreds of French and international sources.

Key features: semantic enrichment (thematic, geographical, temporal) in three languages (FR, EN, ES), multilingual search, integration with HAL, OpenEdition, NAKALA, and DARIAH services, using controlled vocabularies (LCSH, Data BnF, Pactols…).

🔗 https://isidore.science · Documentation

Stylo — Semantic Text Editor

Stylo is a web-based semantic writing tool for academic writing in the humanities. It uses Markdown, YAML, and BibTeX to separate content from formatting, enabling rich export to HTML, PDF, XML-TEI, DOCX, and EPUB.

Originally developed by CRIHN (Université de Montréal) and maintained in partnership with Huma-Num.

🔗 https://stylo.huma-num.fr · Documentation

Huma-Num Box

A cloud storage solution for large-volume, cold or warm data — designed for long-term preservation through replication across multiple sites in France (several terabytes). Distinct from ShareDocs (see below), which targets everyday collaborative file management.

🔗 Documentation · Access via: cogrid@huma-num.fr

Virtual Machines (HN-SSH)

On-demand virtual computing environments for projects requiring custom software stacks, databases, or web applications. Particularly useful for hosting digital editions, project databases, or web services over the long term. Access via project submission to the Comité de la grille.

🔗 Documentation · Access via: cogrid@huma-num.fr

ShareDocs — File Management and Processing

ShareDocs is Huma-Num’s secure online file manager, based on the FileRun application. It is hosted on the French research network (RENATER/CNRS) — entirely within French jurisdiction, unlike commercial cloud services.

It can be used via a web browser, a WebDAV client, or a file synchronisation tool (e.g. NextCloud). Storage capacity goes up to ~1 terabyte per project account. Collaborative document editing is available via OnlyOffice (data stays on Huma-Num servers).

Key uses: storing, organising, and sharing project files (photos, transcriptions, datasets); preparing files before deposit in NAKALA; collaborating with a team.

🔗 https://sharedocs.huma-num.fr · Documentation

Processing Tools — hnTools_watchFolder

One of ShareDocs’ most powerful features is the watchFolder system: a directory within each ShareDocs account that triggers automated processing pipelines when a file is deposited. Results are delivered back into ShareDocs and the user is notified by email.

NoteHow it works

Drop a file into the appropriate subfolder → the system detects it, processes it, and returns the result. Source files are automatically deleted after 21 days.

Tool Function Notes
Abbyy FineReader (OCR) High-quality OCR on images and PDFs Quota-limited; contract valid to June 2026
Tesseract (OCR) Open-source OCR, no quota Many languages supported
Whisper (speech-to-text) Audio/video transcription Models: small / medium / large / with_speaker; data stays on CNRS/IN2P3 servers
FFmpeg (transcoding) Audio and video format conversion Most codecs supported
Tip

The Whisper transcription service is particularly useful for oral history, interview-based, or linguistic research. The with_speaker model adds speaker diarisation. Note that, like all generative AI tools, it can produce errors and hallucinations — always verify transcriptions.

🔗 Documentation — Processing tools

Other Services

Huma-Num provides a range of additional tools for collaboration, project management, data modelling, and text analysis. Most are accessible via HumanID.

GitLab — Code and Script Repository

Huma-Num’s own GitLab instance for hosting, versioning, and sharing code, scripts, and software. A sovereign alternative to GitHub, hosted on French research infrastructure. Public repositories are automatically referenced on code.gouv.fr (DINUM) and archived on Software Heritage (INRIA).

🔗 https://gitlab.huma-num.fr · Documentation

Mattermost — Team Messaging

A self-hosted, open-source instant messaging platform for research teams. Channel-based communication, file sharing, and notifications — all within the Huma-Num infrastructure, without data leaving French servers.

🔗 Documentation

Kanboard — Project Management

A Kanban-based project management tool for organising tasks visually across a team. Supports task cards, deadlines, activity feeds, and access control. Useful for coordinating fieldwork, editorial workflows, or consortium activities.

🔗 Documentation

Heurist — Research Database Builder

Heurist is an open-source tool for building richly structured relational databases without programming, entirely in a web browser. It was designed specifically for humanities and social science research, handling heterogeneous, interconnected entities (people, places, events, objects, sources, concepts).

Originally developed at the University of Sydney (Archaeological Computing Laboratory) by Ian Johnson; Huma-Num hosts a dedicated instance for French SSH research teams.

🔗 https://heurist.huma-num.fr · Documentation

TXM — Textometric Corpus Analysis

TXM is a corpus analysis platform for textometric research — concordances, frequency lists, co-occurrence analysis, factorial correspondence analysis. Developed by UMR IHRIM / ENS de Lyon, it can process large annotated text corpora with a rich web interface.

Huma-Num deploys dedicated TXM portals on virtual machines for research teams who need to host and publish their corpora online. Support and training are provided by IHRIM / ENS de Lyon.

🔗 textometrie.org · Documentation · Requests: cogrid@huma-num.fr

OpenTheso — Thesaurus Management

OpenTheso is an open-source, multilingual thesaurus management system for creating, maintaining, and publishing controlled vocabularies aligned with the SKOS (Simple Knowledge Organization System) standard and linked open data principles. It is used by numerous Huma-Num consortiums and research projects — notably for Pactols, the multilingual thesaurus for ancient Mediterranean studies.

Each term can be assigned a persistent URI, making vocabularies interoperable with other databases and repositories (NAKALA, Heurist, DARIAH services, etc.).

🔗 https://opentheso.huma-num.fr · Documentation

Data Management Plans

For doctoral researchers, a Data Management Plan (DMP) is increasingly required by funding bodies (ANR, ERC, Horizon Europe):

Access for UPPA Researchers

Huma-Num services are accessible to researchers affiliated with CNRS, universities, and research institutions in France. UPPA doctoral students can access most services. The process:

  1. Create a HumanID account at humanid.huma-num.fr
  2. Request access to the specific service(s) needed from your account
  3. For services requiring project validation (virtual machines, TXM portals, Huma-Num Box), contact cogrid@huma-num.fr with a brief project description
  4. For assistance once access is granted: assistance@huma-num.fr