Reading between the lines of big data and data analytics

Friday 19 August 2016

Timos Sellis heads Swinburne’s new Centre for Big Data and Data Analytics.

In summary

This article featured in Swinburne’s new ‘Research Impact’ magazine, produced in association with Nature Publishing Group.

 Whether tracking the movement of shoppers in a mall, or the patterns of stars in the night sky, the new Swinburne University of Technology Centre for Big Data and Data Analytics is set to streamline the sorting of millions of lines of data to generate faster answers. “We want to collaborate with other researchers at Swinburne and external stakeholders, in solving complex data analytics problems,” says Professor Timos Sellis, who has been appointed the centre’s director.

Sellis is an international leader in the field of data science and has an illustrious history of developing new methods for better analysis of non-numerical data. He was the primary inventor, in 1987, of a database structure that improved the storage and retrieval of vast quantities of non-numerical information. The method, known as the R+-Tree, was adopted by industry, and has since been referred to by many companies including IBM, Microsoft and Oracle in patents.

While data science was once purely the domain of finance, it is now applicable across many fields, from health to urban design and the natural sciences.

The new centre will come up with ways to best perform complex analysis on information incorporated from different sources and formats.

“If you ask any company working on an application that uses data, they spend 90 per cent of their time preparing the data and then 10 per cent of the time to actually process the data. First, you have to clean it and make sure that it’s well formatted,” he says.

The centre will also explore ways to quickly and accurately process continuous data feeds from moving sources, such as cars and people.

Sellis is talking with management teams of large shopping complexes about how to process data collected from anonymous shoppers’ phones via their Wi-Fi connection. “We use this information to get an idea of how people move in the shopping mall. This can be useful for security and to understand the behaviour of shoppers — are they there for the supermarket, window shopping or do they mostly come to eat at the food court?”

On campus, Sellis is also looking forward to his centre working with the Swinburne Centre for Astrophysics and Supercomputing to develop new methods to process the vast quantities of astronomical data continuously collected.