Skip to content Search
Search our website:

Optimisation of graph database queries

Optimisation of graph database queries

Principal supervisor: Prof Peter Wood

In recent years graph databases systems, such as Neo4j, have become ever more prevalent and useful. At the same time, a number of query languages for such systems have been proposed and implemented. Graph query languages are usually declarative in nature, and therefore require a sophisticated query optimiser to ensure that query execution times are acceptable.

Most graph database systems do not require a schema to be defined for the data being stored, although a number of schema languages have been proposed. The availability of such schemas provides additional opportunities for optimising queries. This is especially the case when the queries are generated automatically by a so-called flexible querying system, which provides a user with mechanisms to explore graph data by automatically modifying queries in order to provide additional answers.

This project aims to identify the ways in which schema information can be used to optimise graph queries, particularly those generated by a flexible querying system. The investigation is expected to comprise both theoretical and practical aspects, including the implementation or extension of a graph query optimiser.

Candidate Requirements:

Candidates are expected to have a strong background in database systems in general and query languages in particular. Understanding of computational complexity issues would be an advantage. Extensive programming experience is essential.

Key References:

[1] Angela Bonifati, George Fletcher, Hannes Voigt, Nikolay Yakovets: Querying Graphs. Morgan & Claypool Publishers, 2018.

[2] Peter Wood: Query languages for graph databases. SIGMOD Rec. 41(1): 50-60 (2012).

[3] George Fletcher, Alexandra Poulovassilis, Petra Selmer, Peter Wood: Approximate Querying for the Property Graph Language Cypher. IEEE BigData 2019: 617-622.

Further details about the project may be obtained from:

Principal supervisor: Prof Peter Wood

Further information about PhDs at CSIS is available via this link.