Für diese Seite
DigiPolis
Spatial information services prove to be integral part of current and future IT systems and applications. Digital cities, regions, countries and globes are used for communication and decision-making in almost all application areas. Well-known examples are Google Earth or Virtual Earth from Microsoft. The BMBF-funded joint project
DigiPolis develops a 3D information system for regions in the range of less than one square kilometers including indoor areas like e.g. shopping malls, and stations.
DigiPolis is funded by the German Federal Ministry of Education and Research (BMBF) under the support code 03WKP07B.
Self-Organized Semantic Storage Service (S4)
The working group Networked Based Information Systems (AG NBI) develops the underlying storage service for the DigiPolis project. The goal is to design a portable platform that allows distributed storage and retrieval of semantic information. To gain a scalable and adaptive system behaviour we use swarm intelligence concepts leading to a self-organizing system. We build on the following promising approaches:
SwarmLinda
SwarmLinda is an experimental self-organized storage service that emerges from the combination of the Linda tuple space with Swarm algorithms. A tuple space is an abstract, decoupled coordination model which uses tuples (ordered, finite sequences of typed data objects) as its only data structure. Linda is a coordination language which defines simple primitives for storing and retrieving tuples from such a tuple space. Swarm Algorithms are highly scalable algorithms which are derived from the behaviour of individuals in natural swarms. Swarm algorithms solve optimization problems by operating on a two-dimensional model of the environment in a self-organized fashion. They support monolithic as well as distributed execution environments. The primary goal is to organize the stored data in clusters of similar data in order to facilitate quick access.
Semantic Web Technologies
Semantic Web technologies are formalisms and methods for semantic annotation of information and derivation of new knowledge. These are used in the future WWW and in advanced web-based information systems. For our purposes the modelling languages RDF/OWL-Lite and the query language SPARQL are of relevance.
Semantic Tuple Spaces
Semantic tuple spaces, i.e. tuple spaces which store and maintain semantic information are primary research subject of Networked Information Systems research group (AG NBI). Among others a comprehensive theoretical description of Linda tuple spaces for RDF data, called Semantic Web Spaces, was concieved. Furthermore RDFSwarms, a modification of SwarmLinda supporting RDF triples, was developed and tested experimentally. First approaches for clustering mechanisms that order RDF resources regarding to their URIs or concept relations were investigated. These concepts are followed up in this project intensively.
The experimental results from SwarmLinda and semantic tuple spaces are the starting point for the development of the Self-Organized Semantic Storage Service. On the basis of this preliminary work we will design swarm algorithms that cluster RDF data for effective data retrieval. The Web languages RDF / OWL-Lite serve as basis for the system's user interface.
Team & Contact
Project Lead
- Prof. Dr.-Ing. Robert Tolksdorf
Research Staff
- Dipl.-Inf. Hannes Mühleisen
- Dipl.-Inform. Tilman Walther
Student Assistants
- Alexander Dümont
- Sascha Gennrich
- Dennie Kabul
- Kürşad Özarpat
Additional Staff
Contact Information
