Every dataset we release is shared through a lightweight Google Drive workflow — request access, tell us how you intend to use it, and we review and grant within a few working days.
Each set is versioned, citable via DOI, and accompanied by a descriptor on IEEE DataPort. Click a DOI badge to copy it.
A curated MRI imaging set built for self-supervised pre-training and downstream brain tumor classification. Labeled across four diagnostic classes, distributed as a single archive ready for federated learning pipelines.
A multilingual corpus of curated tweets used to train and evaluate models distinguishing extremist content from adjacent but lawful speech. Released to support reproducible ethical-content research.
A larger companion corpus offering multi-label annotations on extremist content in tweet text — designed for fine-grained NLP studies on intersecting categories of harmful speech.
Lightweight by design — we keep it human, not a form-wall.
Pick the dataset that fits your research and grab its DOI from the card above. We'll need this in your email.
Email us at archeintelligencelab@gmail.com with the details listed in the checklist below. Two or three sentences is plenty.
We review within a few working days and grant Drive access to the Google account you provided. Citation guidance follows.
By requesting access, you agree to use our data responsibly — these four points are non-negotiable.
Data may not be re-hosted, mirrored, or redistributed in any form. Send collaborators back to the lab to request their own access.
Cite the dataset using the DOI we provide and acknowledge Arché Intelligence Lab in any work that draws from it — papers, preprints, talks.
Use is limited to non-commercial academic research. Commercial or production use requires a separate written agreement.
Findings must respect the privacy of subjects, the safety of vulnerable groups, and the spirit in which the data was collected.
Send us a short email with your research context. We review within a few working days and share via Google Drive — no form-walls, no waiting lists.