The di.me project researches technologies that enable the use of personal data in a controlled, trustworthy, and intelligent way. The project develops a di.me platform that incorporates user-control deeply in design: a private service – the di.me userware -- offers a central node in a decentralized network, connecting with distinct identities to other users or external services, e.g. social networking platforms. Intelligent features will guide users e.g. by context-aware advice on trust and privacy, or on how to organise the personal information sphere.
This deliverables describes the Data Mining and Semantic Matching engines, two modules behind the Semantic Lifting and Integration objectives of the project. The various techniques employed for the extraction of heterogenous data from distributed sources – of a structured, semi-structured or unstructured natures – are described; together with the method for their representation against the available domain ontologies, and the techniques thereafter applied in order to determine semantic equivalence between two items that refer to the same object. The purpose of this deliverable is to show how these components have been integrated and how they are used within the digital.me userware.
About
- Dissemination level: Public
- Nature: Report
- Workpackage: WP 3 Semantic Representation, Lifting and Processing of Personal Data in Context