Intelligent Information Management / Web Engineering

In Semantic Web Applications, data is typically represented using the Resource Description Framework (RDF) and stored in triplestores, where it can be accessed using the SPARQL query language. While simple SPARQL queries generally perform reasonably well, execution times increase rapidly as the query complexity increases. This can be observed especially in cases where comprehensive information about a specific entity instance with a substantial number of associated attributes is retrieved. This can result in reduced system responsiveness, which in turn may result in degraded user experience.

The Shape Expressions Language (ShEx) provides the means to describe the structural complexity of such data by defining the shape of entities, such as their properties and relationships. The objective of this thesis is to investigate how ShEx schemas can be exploited to improve the performance of read-only queries. First, a requirements analysis has to be conducted, and the state of the art of SPARQL query optimization strategies has to be reviewed. Existing solutions need to be classified and evaluated based on the identified requirements. Based on the findings, a concept for a query execution approach leveraging shape descriptions has to be conceptualized, implemented, and assessed in a suitable evaluation.