LSP.13. Hybrid Database Systems

Relational databases have without doubt become the most common standard of managing and querying information. They are used in a variety of contexts with many different requirements including analytical querying and business intelligence. However, with the growing popularity of the internet and social networks, new requirements have emerged and graph-structured data is becoming more and more important. To address the characteristics of graph data more naturally, native graph databases have been developed.

Both relational databases and graph databases have strengths and weaknesses depending on the characteristics of the data and the types of queries that need to be evaluated. For instance, finding the shortest path between two particular nodes can be computed more naturally in a graph database whereas finding a set of entities based on some attribute values is more natural for relational databases.

The aim of this topic is therefore to develop a hybrid system that unifies the strengths of both relational databases and graph databases by analyzing the strengths and weaknesses in detail, designing a suitable storage layer, and developing appropriate query optimization techniques for analytical queries. As the expressivity of relational and graph query languages differs, an appropriate query language needs to be defined or an extension to an existing one.

Main Advisor at Technische Universität Dresden (TUD)
Co‐advisor at Aalborg Universitet (AAU)