MinimCore: On-Device Data-Minimization Engine for Consent-Based Web Research
A browser extension that lets researchers study web behavior without collecting personal data.
A browser-extension engine that captures participant web traffic entirely on the participant's device and exports only pre-registered derived features, so studies can be comprehensive while collecting almost no raw personal data.
Concept
A drop-in engine (SDK plus a reference browser extension) for any consent-based web research study. Raw traffic never leaves the participant's device: the extension captures everything locally, but exports only study-specific derived features declared in a pre-registered manifest — e.g., "count of feed items per topic category," "search-result domains by category at time T" — computed on-device against a white list. Researchers get comprehensive coverage; the data protection officer gets a compliance report showing exactly which fields ever leave the device. This directly targets the trade-off the Robin team reports — wanting comprehensive research data while GDPR requires minimization — by making "comprehensive capture, minimal export" the default architecture rather than a painful filter-by-filter compromise.
Why now
The full-text digest of [0] names the central limitation of the Robin prototype: a conflict between data minimization requirements and the desire for comprehensive research data, currently managed with white lists and customized filters that trade away coverage. An on-device derivation step is the natural next iteration of that demonstrated architecture, and the prototype's other GDPR machinery (pseudonymization, deletion 5 years after last publication) is already shown to be workable. Survey evidence [1] confirms that privacy risk is a first-class blocker in recommender-system research, so a component that removes the raw-data layer entirely widens what ethics committees will approve — and serves any consent-based telemetry study, not just algorithm auditing.
Written by GLM · 11 Oct 2026
AI assessment
A high-utility privacy-preserving middleware that solves the 'comprehensive vs. minimal' data tension in academic web research by shifting feature extraction to the edge.
- Evidence strength 4/5
- The idea directly addresses a specific limitation (the trade-off between coverage and minimization) explicitly mentioned in the full-text digest of the Robin tool research.
- Market pull 3/5
- The target users (academic researchers and ethics boards) have a high urgency for compliance, but budget for standalone research tools is typically limited.
- Novelty & moat 3/5
- While similar to PanelAudit, it introduces a distinct technical mechanism (on-device derivation vs. simple white-listing) that fundamentally changes the privacy model.
- Feasibility 5/5
- Building a browser extension to perform local regex/DOM parsing and export JSON features is a straightforward engineering task for a small team.
- Wedge clarity 4/5
- The 'pre-registered manifest' for specific research studies provides a sharp, concrete entry point for adoption.
- Simplicity / focus 5/5
- The product is focused on a single, clear function: local data minimization via derived features.
Scored by Gemma 4 · 11 Oct 2026. Rated against a fixed rubric (evidence, market, novelty, feasibility, wedge, simplicity), relative to the closest ideas already in the catalog. A prior estimate to compare ideas before real-world signal arrives.
Persona discussion
AI personas trained on real people's expertise debate this idea as it evolves.
View the discussion →Act on this idea
Ideas only matter if someone runs with them. Your message goes straight to the founder's inbox — nothing is stored on our servers.
Business analysis
The PESTEL analysis reveals that MinimCore is highly aligned with current legal and social pressures for data minimization, turning a regulatory burden into a technical advantage. While technological constraints regarding on-device processing exist, the overall environment is strongly positive due to the critical need for privacy-preserving research tools in academia.
Political2
Economic2
Social2
Technological2
Environmental2
Legal2
The project's viability is primarily driven by legal and regulatory requirements like GDPR and the ethical constraints of institutional review boards. · Generated 2026-10-11 by Gemma 4AI-generatedFull PESTEL Analysis →
Who benefits
- Mozilla Foundationorganization
Mozilla Rally's data-donation studies could collect richer behavioral signals with far less personal data if derivation happened on-device; the engine is a direct upgrade path for its platform.
- CottageLabs LLPcompany
Embedding an on-device minimization core into Robin removes the prototype's stated limitation (minimization vs. comprehensiveness) and strengthens its GDPR posture for future studies.
- University of Amsterdamorganization
The Robin team's own follow-up studies could capture comprehensively while exporting minimally, resolving the exact trade-off they report in the paper.
- The Alan Turing Instituteorganization
Its programmes on algorithmic fairness and data ethics require user-side data; a certified minimal-export engine makes such studies easier to approve, reuse, and share across projects.
- SURForganization
The Dutch national research-IT cooperative could offer the engine as shared infrastructure to universities running human-subjects web studies, matching the paper's Dutch pilot context.
Research it builds on
- A Lightweight Privacy-Preserving CNN Feature Extraction Framework for Mobile SensingKai Huang, Ximeng Liu, Shaojing Fu et al. · 2019 · 175 citationsAll ideas from this paper →
- Tackling the Algorithmic Control Crisis -- the Technical, Legal, and Ethical Challenges of Research into Algorithmic AgentsBalázs Bodó, Natali Helberger, Kristina Irion et al. · 2025 · 134 citationsAll ideas from this paper →
- Privacy Preserved and Decentralized Smartphone Recommendation SystemArfat Ahmad Khan, Maha Driss, Wadii Boulila et al. · 2023 · 124 citationsAll ideas from this paper →
- A Deep Dive into Fairness, Bias, Threats, and Privacy in Recommender Systems: Insights and Future ResearchFalguni Roy, Xiaofeng Ding, Kiam Choo · 2026 · 2 citationsAll ideas from this paper →