The DataVaults Consortium worked on the definition of the integrated DataVaults Methodology and development of a set of high-level operation scenarios that help to identify the expected behaviour of the DataVaults platform and led to the formulation of the DataVaults Most Valuable Product (MVP).
The results of this work are reported in D1.3 – DataVaults MVP and Usage Scenarios.
The DataVaults Methodology
The initial DataVaults Methodology is structured in eight interrelated Phases, that capture the overall scope of DataVaults:
- Data Retrieval
- Data Transformation & Enrichment
- Asset Storage at Individual’s Side
- Asset Sharing to DataVaults Cloud Platform
- DataVaults Cloud Platform Asset Storage
- Asset Exploration & Extraction,
- Data Analytics
- Added Value Services
Each Phase has its own Operations that outline at a high level the distinctive tasks that will be performed by the platform (Figure 1).
The different needs of users from DataVaults are illustrated in eleven high level platform operation scenarios, that guide the reader through the envisioned functionalities. The scenarios are divided in three groups, around the main involved Actors (i.e. driven by the Individual, the Data Seeker, or the DataVaults Data Scientist). However, they include not only the scenario descriptions, sequence of steps and workflow diagrams, but investigate ethical, security and GDPR-related aspects, in order to ensure compliance, highlight points that shall be researched in depth and pinpoint functionalities that are required for achieving maximum legal compliance, privacy and security.
Feature Engineering and DataVaults MVP
Based on the operation scenarios and the DataVaults Methodology, a set of 79 features has been extracted. The value and implementation complexity of these features are parameters that play a very important role in deciding what will be implemented and when, both from a technical and a business aspect. For this reason, the consortium assessed all features, based on their expertise, and ended up with the following graphical distribution of features in four areas, representing complexity and value. The features are also labelled based on their expected added business value for the five demonstrators, using the MoSCoW method (Featured Image).
This prioritised list of features leads to the first DataVaults MVP (Most Valuable Product); a product that delivers the maximum value to the involved stakeholders (demonstrators and technical partners) both during and post project implementation, given any technical, time or other constraints. The MVP will be the basis that will facilitate the architecture design and guide the development activities for the various components that comprise the DataVaults Personal Data Platform.