This prize awards the VLDB paper from 10 years ago that has had the most impact. Boncz was honoured for the influential paper ‘How Good Are Query Optimizers, Really?’, co-authored in 2015 with Viktor Leis, Andrey Gubichev, Atanas Mirchev, Alfons Kemper, and Thomas Neumann from TU Munich.
Crucial challenge
The award-winning article identified the cardinality estimation technique as the crucial challenge in query optimization. When users search in a large dataset, the database system must decide in which order it should best bring together different pieces of information – much like choosing the right sequence of steps in solving a puzzle. If the order is wrong, queries may run unnecessarily slowly. The article revealed that many popular systems were weaker at this task than expected. The paper also introduced the Join Order Benchmark (JOB), now a global standard for testing query optimizers.