A research project to help ‘see’ huge amounts of extremely complex data will be led by a senior computer science researcher at Charles Sturt University (CSU).

 

CSU’s Professor Junbin Gao recently won a $210 000 research grant from the prestigious Australian Research Council to investigate new and better ways of managing, analysing and visualising complex and ‘big’ data used in modern business and scientific fields.

 

“We are aiming to develop a framework that uses human sight to reduce the complexity of data which is generated in everyday human activities on such websites as Facebook, ebay, and Youtube,” said Professor Gao, who is a research leader with the University’s Centre for Research in Complex Systems.

 

 “The ‘curse of dimensionality’ is a nightmare for most paradigms by which computers ‘learn’. It affects most algorithms used in computers and mobile phones and has strange affects on data that is represented in higher dimensions. We are used to ‘seeing’ data analysed in two or three dimensional graphs, but most people have trouble with perceiving data in four or more dimensions.

 

“People like to ‘see’ through the complex data and to make sense of it for such purposes as business analysis that identifies trends or patterns of customers from extensive and changing data collected daily.

 

“We want to simplify this huge amount of data and reduce their dimensions while retaining the most sensible information contained in the data for analysis.”

Researchers from the University of Reading in the UK will also be part of the team led by Professor Gao which will develop in the next three years a systematic approach that designs and analyses basic algorithms to reduce dimensionality. He has worked on the topic for the past five years and established a national and international reputation in the area.

 

Professor Gao believes this research will have tremendous applications in the modern world. “In our research we will also explore applications to analyse business data and digital images. We also see potential application in the next generation of printers, in drug design through visualizing the protein folding process, and in object tracking systems used for national border protection.”

 

Students in CSU’s Bachelor of Computer Science (Games Technology) course will also be involved in the research project.