When Michael Stewart moved from Northam in country WA to Perth to study Computer Science and Software Engineering at UWA, he had no idea that together with his senior colleagues, he would play a major part in revolutionising data technology that can benefit many industries and communities.
Early in 2016, Michael and his PhD supervisors – Dr Wei Liu and Associate Professor Rachel Cardell-Oliver – were approached by Professor Mark Griffin (then at UWA, now leads Curtin Future of Work Institute), an expert in workplace psychology, to devise a new way for discovering causes behind accidents and injuries in the mining industry.
Michael and his supervisors quickly realised that many businesses and organisations in various industries struggle to find a simple yet highly effective way of analysing data.
“In general, data is simply another word for information. But in computing and business, data refers to information that is machine-readable as opposed to human-readable,” explained Michael.
“Processing data in their entirety can give us a far more complete picture of the world around us, so being able to accurately and easily analyse data to inform decisions is a powerful tool for any industry.”
What Michael and his supervisors also discovered was a research gap when it comes to an easy but effective and accurate way of reading and interpreting data.
“The process of analysing written reports, such as workplace accidents, is typically performed manually by a team of experts. It requires a considerable amount of time and effort,” explained Michael.
So together they worked to come up with a solution – Aquila – a tool for analysing unstructured data. It features a simple and intuitive web-based interface that allows for the discovery of latent knowledge buried within textual data.
Understanding that complexity is a barrier for accessing data analysis, Aquila has been developed to be used by anyone with no programming experience.
“The key innovation of Aquila is its incorporation of numerous natural language processing
(NLP) and data mining techniques,” said Michael.
“In short, what Aquila does is simplify the interaction between human language and computers, to provide insights and patterns that we can then use to make better decisions in the workplace, in business, in whatever industry.”
“It is, to the best of our knowledge, the first ever dedicated system to computationally analyse short text log data, and the first system to combine NLP and data mining in this manner.”
The potential for Aquila to be used to improve outcomes for communities and businesses is huge!
“Many companies across a wide range of industries possess large volumes of unstructured data – for example, doctor’s reports, traffic reports, police reports and maintenance records.” said Michael.
“Valuable insights can be drawn from those data sets, and the knowledge then used for driving improvement such as better traffic management, timely repairs of equipment to reduce wear and tear, suburbs that require higher safety networks and more.”
“There was considerable research value in an accessible text-to-knowledge system with no programming experience necessary to use it.” Michael said.
The untapped potential for Aquila’s innovation has been recognised across the State with the technology winning the Research and Innovation Project of the Year at the 28th WAITTA INCITE Awards – WA’s premier event for recognising and celebrating innovators in the Information and Communications Technology (ICT) industry.
Improvements and new features in the coming months will see Aquila incorporate deep learning algorithms and knowledge graph visualisations. Michael also spoke of the project’s longer-term goals.
“Our primary long term goal is to introduce a question-answering system, where users may ask questions about their data directly in plain English and get a relevant and insightful answer,” Michael explained.
“There are so many exciting technologies being developed right here in WA. It has been fantastic to share Aquila with the Perth tech industry and to see its potential to significantly impact on how businesses and government agencies operate. We are looking forward to sharing it further afield.”
See Aquila online: http://agent.csse.uwa.edu.au/aquila/