Decentralized, Self-organizing methods for efficient data access
|
Methods addressing the issue of decentralized, self-organizing, efficient data access are based on scalable data access structures as a basic infrastructure for efficiently locating data objects without relying on central directories and mobility of data objects, where data is actively locating and negotiating with nodes where it is most needed.
More Information here
|
Global semantic interoperability
|
Here we will analyze local metadata annotation schemes and networks of user-supplied metadata translations (local semantic agreements) in order to detect globally agreed semantics through reflection, that then can be used for answering structured requests using gossiping and metadata translation.
More Information here
|
For programming support we develop an architecture and implementation for mobile code, based on the functional and object-oriented Scala language. The ability to transfer code (or, more precisely, closures) between sites increases the potential for flexible self-organization, but can pose security problems. We will address these problems through a type-systematic approach. One important application of this work will be in expressing the filter functions in the type-based publish/subscribe framework.
More Information here
|
For communication support in programming we develop type-based publish/subscribe mechanisms to interoperable type-based pub/sub across different data models and programming languages. Type-based Publish-Sub-scribe (TPS) can be viewed as a high-level candidate OS abstraction for building Peer-to-Peer (P2P) applications over WANs. TPS has a completely decoupled underlying communication scheme that perfectly fits self-organizing systems, while ensuring object encapsulation and type safety.
More Information here
|
Spatial Semantic Services for Mobile Users
|
Here we will develop the location-based spatial information services for the mobile users by designing and implementing semantic extraction services on the basis of mobile ad-hoc-networks infrastructure and specifications. Our focus is to design facilities to address ontological issues, semantic extraction from structured and unstructured sources, multi-modal user interface, as well as personalization and matching of preformatted queries. The targeted result is to provide users with a consistent and desirable view of the locally available information.
More Information here
|
Contracts in Terminode Networks
|
This research addresses the modeling of contracts in the context of Terminode networks. A contract is a mechanism used to publish and invoke what systems can do. Contracts are central to the organization of Terminode networks. This research aims at defining a method to define and reason about contracts. Our edge is to base this method on system sciences and on our ontology for general system modeling. We also combine the experience coming from contracts on software engineering (for the precision of the description) and in agents - AI (for the capability to reason and plan activities).
More Information here
|
|
