Data analysis mines information

Information definitely exists but how are we to make sense of it? Contemporary science has at its disposal ever more efficient methods for measuring and information retrieval. Computers are capable of increasingly more robust calculations. The essential content can be gleaned from ever expanding materials with data processing methods such as those being developed in the Centre of Excellence in Algorithmic Data Analysis Research.

“When you are trying to determine what kind of genetic malfunction is the cause of a disease, you have to process the data on the simultaneous functioning of over 20,000 genes,” explains the Director of the Centre of Excellence, Professor Esko Ukkonen, giving a typical data analysis problem.

Another type of enormous mass of data is battled with when one tries to trace how new things appear on the Internet. With automatic analysis, linguists can draw dialect boundaries on a map, while zoologists can present areas of distribution of various species.

 The Centre of Excellence develops data analysis methods in close co-operation with, for example, researchers in molecular biology, medicine, environmental research, linguistics and telecommunications. Analysing methods, that is algorithms, are being developed for fresh problems in these fields and, on the basis of them, computer programs are being created to deal with the enormous amounts of information.

“Core areas in the development are combinatorial pattern matching and data mining. We are the leading research group in the world in these fields,” says Ukkonen.

Both methods seek various regularities and interesting phenomena from extensive collections of data and such character strings as natural languages or DNA.

The Centre of Excellence is about to start its second operating period. A new feature is Adjunct Professor Aapo Hyvärinen’s research group which studies hidden structures in very complex data.
”This method is needed for example in modelling the optical system. Vision gives us a huge amount of data, of which only the significant features are distinguished,” says Ukkonen.

The Centre, which has a staff of 70, is active at the University of Helsinki and the Helsinki University of Technology.

Centre of Excellence in Algorithmic Data Analysis Research

Text: Tiina Männistö
Photo: Veikko Somerpuro
Translation: Valtasana Oy