Research

Some research projects are described below. For more details take a look at DASlab.

Soon everyone is going to be a "data-scientist". dbTouch explores a new direction towards database systems that are driven by touch interfaces through intuitive gestures as opposed to using SQL.The aim is to bring the power of data analytics to everyone with access to a standard touch device. The main challenge is in designing a new class of database kernels that can be responsive to touch in order to provide interactive exploration. 

Before using a database system, we need to tune it with the proper indexes. In turn, this requires workload knowledge and idle time. However, both of these are scarce resources in the big data era. In addition, having to tune a database system, means that one has to be a database expert to do so. Adaptive indexing aims to remove the bottleneck of indexing in database systems, allowing for quick access to new data and also allowing one to use a database system without the need for an expert database administrator. 

As data sizes grow, even loading the data inside a database system becomes a major bottleneck. Often, we simply want to take a quick look at the data to figure out which parts are interesting and what kind of queries we would like to do. Here, we aim to provide database-like functionality on top of raw data eliminating the bottleneck of data loading but still maintaining the rich query processing features found in database systems. 

As hardware and data-driven applications evolve, there is an increasing need to reconsider database architectures. We study modern systems and the various options regarding how we should store and access data.