Trisolda Jakub Yaghob Charles University in Prague, Czech Rep.

Trisolda overview Import paths SemWeb repository Query environments Execution environment

Trisolda repository Stores incoming data Retrieves results for queries Stores used ontology Import interface, query interface, application server Trisolda datastore – holds data in any format

Trisolda application server Not all accurate data and inferred knowledge available at the moment of data import The knowledge of the world is not accurate Background worker Inferencing, data unifications, reasoner Uses import and query APIs Framework for server’s plug-ins Other experimental implementations of reasoners, unifiers, etc.

Import paths Direct import data in data sources converters to the used ontology Crawling wild Web Egothor web crawler parsed pages stored deductors deduce data and its ontology Batch insert Immediate insert

Query API Based on simple graph matching and relational algebra Simple graph matching – SPARQL-like Query: set of RDF triples with variables Result: multiset of possible variable mapping – a relation (table) (the good old) relational algebra Not another SQL-like language set of C++ classes and operators query API used by software Query evaluation different level of support by storage engines

Query environments Query Environments present outputs from the repository Using Query API Examples: SPARQL compiler, Tequila compiler, repository browser, RDF visualizer, semantic executors, …

Conductors & executors Executor Atomic unit of execution No intelligence Interconnects both virtual and real worlds Conductor Realizes business process Unit of semantic intelligence Composes and controls executors

Conclusions & future work Conclusions Working infrastructure for gathering, storing and querying semantic data Platform for future semantic web research Future work Simpler web crawler Native distributed parallel RDF storage NLP?