Fast and Scalable Recommendation Computation in Semistructured Collaborative Information Systems

SnoopyDB is a concept and also prototype of an information system, which enable input of any information in form of key/value-pairs without requiring any technical knowledge. Significant difference to other such systems is the guidance-system working on already existent data sets for inserting or querying data. This part of the system can snoop additional information from the user during insertion process.
Goal of this master thesis is the further development of the existent prototype. Specifications for interfaces and a modular system has to be developed and implemented. This will enable an easy activation and serial configuration of single modules (e.g. a thesaurus). By improving the algorithms regarding the guidance-system and storage of data, the system should be able to handle larger amounts of data. In order to guarantee full functionality and efficiency, RDF (storage) and SPARQL (interfaces and algorithms) will be used.