"By using brain-inspired techniques, computers can now learn from experience."

Every day, global digital activity increases in speed, volume and diversity. As a result, this demands that computing systems do more than simply calculate an answer – they need to sense, interact and predict.

Associate Professor Damminda Alahakoon, director of the newly-launched Deakin Cognitive Analytics Laboratory (DCAL), says that 'cognitive analytics presents a paradigm shift' because it draws inspiration from the way in which the human brain processes information.

'Cognitive analytics uses "brain-like" thinking to capture and process inputs in parallel, learn high-level concepts, acquire knowledge and draw out informed decisions,' he says.

With almost a decade of global experience in the IT and finance industries, Dr Alahakoon specialises in machine learning, text analytics, data mining and business intelligence and his strong industry background enables him to focus on practical outcomes.

He says that DCAL's technology is based on research into the frontiers of artificial intelligence, cognitive science, computational neuroscience and cognitive psychology.

'Our research focuses on the search of computationally-plausible models, techniques and methods that can be adapted to enhance and extend current practices in data analytics. By using brain-inspired techniques, computers can now learn from experience by unscrambling complex data to recognise associations,' he explains.

Unlike traditional analysis – which depends on structured rules and questions – cognitive analytics thrives on large amounts of data:  the more a machine learns, the greater quality of insight it delivers.

In the current information-intensive environment, Dr Alahakoon says the possibilities for cognitive analytics are boundless and DCAL's focus is on developing technology and software whilst also providing expertise to a range of clients who require analysis of large quantities of data.

'DCAL has developed a number of new "brain-inspired" computing techniques which make up the basic building blocks for innovative end-to-end analytics solutions. The key building blocks consist of advances in text analytics, incremental learning and real-time data mining, multi-source data integration and sequence discovery. High performance computing techniques using Hadoop supplement the base building blocks in big data environments,' he says.

Established in 2013, DCAL's projects focus on 'real-world' applications which include database and data warehouse development, business intelligence system development, health care data mining, text analytics in horizon scanning, automating essay and short-answer exam marking, smart electricity meter data analytics and intelligent web-based system development.

'With clients in academic, research and commercial sector organisations, we are working on a range of key projects. For example, at Monash Health we're working with radiologist on text analytics techniques that are being used to develop an electronic radiology library while also helping in mining radiology reports and images for better medical decisions,' says Dr Alahakoon.

Industry collaborations form the core of DCAL's work and present the researchers with 'almost infinite' possibilities says Dr Alahakoon.

'In another partnership, we're working on automated exam marking which is now being trialled with the International English Language Testing System. We've teamed up with the Melbourne technology firm Genix Ventures to test this with existing clients.  DCAL's basic technology building blocks could be put together to build solutions for diverse applications in many domains and this has enabled us to develop exciting partnerships and collaborations,' he explains.

The DCAL researchers are also working with RMIT civil engineers and VicRoads on a project that is developing tools to predict the deterioration in road bridges and improve the decision-making process in repair and maintenance.

And in a recent collaboration with Deakin's School of Psychology, DCAL will soon provide assistance in the handling of big data, data integration and incremental learning techniques which extract patterns from brain imaging and patient behaviour.

The demand for the technology and expertise behind these 'real world' solutions is now driving the need for more skilled cognitive analytics professionals says Dr Alahakoon.

'There's a big push to develop analytics expertise, not only in research but also in teaching; Deakin Business School now has a new Master of Business Analytics program and in the future, DCAL plans to introduce new teaching tools based on research projects for both undergraduate and postgraduate students.'