Researchers at the IT University of Copenhagen are developing a new method to analyze extremely large data volumes. The project will help geneticists increase the milk production of Danish cows and has received a grant from The Danish Council for Independent Research.

How can you use complex mathematical algorithms to make milk cheaper? A researcher at the IT University of Copenhagen can answer that question. The 35-year-old scientist has discovered a new way to make the algorithms, i.e. the mathematical recipes which are the backbone of the software used to analyze very large volumes of data.   

Similar to making polls

Rasmus Pagh is heading the project Data Massive Mining by Sampling, which has been granted 5.6 million DDK from Sapere Aude, the career program of The Danish Council for Independent Research. The results of the research can turn out to be an invaluable help for companies who want to find patterns in several millions of information about what their customers search for and what they like to buy. Today it requires very large and expensive IT capabilities to handle so much material, which often makes it difficult to make analysis and statistics.

- Without good algorithms you can experience that the computers will need several thousand years to analyze these large data volumes, and therefore it is useless to us, says Rasmus Pagh.   

- My method is similar to the way you make polls. You use selected information to draw the big picture of people’s opinion.

More Milk

The perspectives of Rasmus Pagh’s research are various. The research team at the IT University of Copenhagen cooperates with geneticists at the Aarhus University, who want to survey 800 million genetic data from cows in order to improve breeding in agriculture.  

- My research will make it easier for us to survey an enormous amount of data.  In this case you will - from this enormous amount of data - find some genetic correlations among breeding cattle, so you can get healthier cows that produce more milk. This will improve the competitiveness of agriculture, which ultimately can lead to cheaper milk for the consumers, says Rasmus Pagh.  

 

 

About data mining

Large volumes of data often contain a hidden, but interesting, knowledge which is inaccessible to us as human beings, because our brain cannot compare millions of information. For example, a bank could get a fairly accurate picture of their clients' hobbies, food habits, and preferred means of transport by comparing customers' bank statement of accounts. Such a project, however, will usually be impossible for bank analysts - unless they know about data mining or business intelligence; methods to analyze extremely large volumes of data.   

Data mining is about to become an invisible but indispensable part of our society. Data mining, for example, improves a doctor’s possibilities to make diagnoses based on information on millions of patients' health. And when your online bookstore recommends the book that you never knew you wanted to read, it is because data mining makes it possible to analyze your clicks and compare them with millions of other customers' clicks. Actually it’s doubtful whether the Oresund Bridge would have been built if data mining had not made it possible to make a precise prediction of the water streams in Oresund from an immeasurable number of log data about the fairway, which has been collected since the 18^th century. However, the potential of data mining is not yet utilized, because it requires a huge IT capacity to handle millions of data. This is why associate professor at the IT University of Copenhagen, Rasmus Pagh, is developing a new method for creating the algorithms, i.e. the backbone of the software, used to analyze and compare large amounts of data. This method will presumably improve opportunities for the use of data-mining.

 

About the IT University
The IT University of Copenhagen has existed since 1999 and is an independent university dedicated to the digital world. There are about 2,000 students and 270 employees here. The IT University researches and teaches in a wide spectrum of topics within the field of IT.   

About Rasmus Pagh
Rasmus Pagh is 35 years old and has received a grant of 5.6 million DDK from the prestigious Sapere Aude,  The Danish Council for Independent Research ´s career program, reserved particularly for talented young scientists.

Additional information
Research Communicator Jari Kickbusch, 72185042, jark@itu.dk

Associate Professor Rasmus Pagh, pagh@itu.dk (At present Rasmus Pagh is in the U.S., but an interview by phone can be arranged)