Imagine trying to find a needle in a haystack, but the haystack is growing every second and you aren't sure what the needle looks like. Instead of letting a computer guess blindly, this system lets a human expert guide the search by giving feedback on the results. It's like having a smart assistant that learns exactly what you care about to give you the most precise answer possible.
Current big data and AutoML tools often ignore the expertise of the actual business user, leading to results that are technically accurate but practically useless or untrustworthy for critical decision-making.
["A secure distributed management system for datasets and knowledge assets with contextual authorization.", "Automated tools for dataset selection, feature augmentation, and data cleaning to ensure data is fit-for-purpose."]
If you are a SOC provider dealing with massive streams of threat data — this project developed an experimentation engine that allows analysts to refine detection models based on their expertise. This ensures alerts are precise and trustworthy rather than overwhelming the team with false positives.
If you are a maintenance provider dealing with heterogeneous sensor data from old and new machinery — this project developed analysis-aware data integration that helps users select the best datasets for the task. This reduces the time spent cleaning data and increases the accuracy of failure predictions.
If you are a mobility consultant dealing with high-speed traffic and public safety data — this project developed a decision support system that integrates simulations and visualizations. This allows city planners to test different traffic variants and choose the one that best meets their specific constraints.
Based on available project data, no commercial pricing or licensing costs are provided as this is an EU-funded research project.
The project is designed to handle extreme data characteristics including volume and speed, and it will be validated across 5 pilot demonstrators to prove its scalability.
Based on available project data, specific IP and licensing terms are not listed; however, the consortium includes 21 partners across 10 countries.
It provides core services for secure distributed management of datasets and automated dataset selection strategies to fit the current analytics task.
The project period runs from 2023-01-01 to 2026-02-28, suggesting that final results will be available by early 2026.
The consortium is well-balanced for commercial transition, featuring a 38% industry ratio with 8 industrial partners, including 5 SMEs. The collaboration of 21 partners across 10 countries suggests a broad market validation strategy, combining the academic rigor of 6 universities and 7 research centers with practical industrial application.
Contact ATHINA-EREVNITIKO KENTRO KAINOTOMIAS in Greece
Contact us to explore licensing opportunities for the experimentation engine.
